Content uploaded by Darren Halpin
Author content
All content in this area was uploaded by Darren Halpin on Jan 12, 2018
Content may be subject to copyright.
Defining Interests: Disambiguation and
the Need for New Distinctions?
Grant Jordan, Darren Halpin and William Maloney
This article notes the systemic lack of conceptual clarity in the social sciences and attempts to illus-
trate the adverse consequences by closer examination of the particular example of the interest group
field. It indicates the significant ambiguities implicit in the term. Not all policy-influencing organ-
isations are interest groups as normally understood, but because there is a lack of an appropriate
label the term interest group is used by default. The article seeks to distinguish between interest
groups and other policy relevant bodies—often corporations or institutions. It finds disadvantages
in adopting a functional interpretation of the interest group term (i.e. any organisation trying to
influence public policy). While the wider range of organisations are crucial in understanding the
making of public policy, it is confusing to assume that this wider population are all interest groups.
The article instead advances the complementary notions of pressure participant, policy participant
and interest group. This slightly expanded repertoire of terms avoids conflating important distinc-
tions, and, in Sartori’s term permits ‘disambiguation’. The core assumption is that the search for
comparative data and exploration of normative questions implies some harmonisation in the inter-
est group currency.
With few—if any—exceptions, concepts in the social sciences are poorly defined:
indeed most prove popular precisely because they have an imprecision that allows
promiscuous application.1Though the press and non-specialist political scientists
think they know when an organisation is, or is not, an interest group, this is an
area where more careful scrutiny produces less rather than more confidence. This
article ‘tests’ a very basic ‘unit of analysis’ crucial to political science, the interest
group, and finds that the elasticity of understanding among scholars in the area
makes cumulative studies (unnecessarily) difficult. In the absence of definitional
clarity, general conclusions about what sorts of groups dominate the democratic
system—an important dimension to interest group study—are almost impossible
to draw with any accuracy. The core proposition in this article2is that different
authors cover different types of organisation in the field loosely delineated by the
interest/pressure group label. While there is value in Karl Popper’s concern that
we should not be ‘goaded into taking seriously words and their meanings’ (Popper
1976, quoted in Gerring 1999, 360), if there are no agreed language-tools there
can be no comparison of conclusions.
The justification for indulging in definitional fine print is that such issues need to
be resolved. W. J. M. Mackenzie described British interest group research in 1955
as ‘wrapped in a haze of common knowledge’ (quoted by Richardson 1999, 181).
The ad hoc practices currently evident in the literature have been a contributory
factor leading to the marginalisation of interest group studies within the political
BJPIR: 2004 VOL 6, 195–212
© Political Studies Association, 2004. Published by Blackwell Publishing Ltd, 9600 Garsington Road,
Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
196 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
science discipline. Frank Baumgartner and Beth Leech (1998, xvii) observed how
researchers had come to adopt definitions based on data availability or to invent
their own definitions to suit particular case studies. Both these practices make it
difficult to weave case-study material together into a coherent picture of the inter-
est group system. Definitions robust enough to support data collection and com-
parison are required: until researchers start counting the same things when they
say they are counting interest groups, conclusions are founded on shifting sands.
This contribution leaves aside the related question as to whether the label ‘pres-
sure group’, ‘interest group’, ‘organised group’ or whatever is the best signal of the
‘universe’. Sometimes these terms are used as synonyms; sometimes different
meanings are imputed to each. The priority in this article is to identify the general
field and to introduce a degree of clarity.3The definitional boundary problem in
the interest group field is most commonly discussed in terms of making distinc-
tions from parties (see definitions cited by Baggott 1995, 2–3), but more common
areas of ‘blurring’ are between interest groups and business or governmental
actors.4While the pressure group sub-discipline has itself generated terms with
remarkable promiscuity (Kimber and Richardson 1973, 1, counted 21), this article
does not try to referee the sub-classification area.
Indeed the chaos that Richard Kimber and Jeremy Richardson reported has been
reinforced by newer generations of terms such as non-governmental organisations
(NGOs) and social movements (SMs).5The latter perhaps signals less formal organ-
isations than the term group but the pressure group literature traditionally included
‘unorganised’ groups. The distinction between organised =interest group and
informal =social movement is questionable. Like the use of NGO, the social move-
ment term seems to suggest a normative approval of the ends of the group rather
than denoting a different type of social organisation. While such terms embody
important distinctions within particular theoretical frameworks, the groups ges-
tured to by theorists as fulfilling a ‘role’ or ‘function’ as an NGO or SM, with their
implied impact in terms of democratisation and representation, are a mixed bag of
organisational animals. In the absence of an agreed terminology for distinguishing
different types, the normative claims for the value of organisations fulfilling the
functions ascribed to NGOs and SMs are generalised to a broader range of bodies
where the case is far weaker. Describing by implied function, rather than by organ-
isational characteristics, obscures important differences. The interest group termi-
nology needs to be fine-tuned to more clearly identify the animals in the interest
group zoo. A more precise definition of interest group-type structures opens up,
but does not replace, the normative debate. A stable set of definitions to separate
out different types of body currently all referred to as interest groups can co-exist
with, even complement, the concepts of SMs and NGOs.
This article is essentially in three parts. The first indicates that this discussion is a
specific case of a wider phenomenon of conceptual imprecision. The second part
tries to demonstrate that there are serious research consequences in the interest-
group area stemming from definitional issues. The final, more contentious, section
offers a ‘solution’ to the problem identified. Inevitably, an attempt to resolve con-
fusion by introducing new terms may be read as adding to the confusion. But the
article assumes this risk is worthwhile.
DEFINING INTERESTS 197
The General Problem of Conceptual Imprecision
While some scholars are impatient at the idea of seeking a coherent and consis-
tent language of inquiry, there is a serious academic tradition that would be sur-
prised at the assumption that the sort of hotchpotch of hand-me-down concepts
with which we try to investigate the social sciences could be ‘right’. Thus Giovanni
Sartori (1984, 15) starts his own analysis by quoting Robert Merton (1958, 114):
‘a good part of the work called “theorising” is taken up with the clarification of
concepts—and rightly so’. Merton was perhaps optimistic. Weber (1949, 105, cited
in Gerring 1999, 359) famously argued that it was in the perpetual reconstruction
of concepts through which we seek to comprehend reality:
The greatest advances in the sphere of the social sciences are substan-
tively tied up with the shift in practical cultural problems and take the
guise of a critique of concept construction.
The article nibbles at the edge of a problem whose core is addressable only through
full engagement with the philosophy of language. For example there is, as set out
by Sartori (1984, 57), Wittgenstein’s idea that ‘the meaning of the word is its use
in the language’. Sartori’s proposition that whatever ‘science’ is, a preliminary con-
dition resides in the formulation of a special and specialised language with distinctive
characteristics precisely to correct the defects of ordinary language, is powerful. He
says, ‘the various sciences ... took off by inventing neologisms [their own techni-
cal vocabulary], by reducing by definition the ambiguity of their key terms, and
by consistently abiding by syntactical rules’. Sartori continues, ‘but a soft social
science that bows to Wittgenstein’s dictum negates the very possibility of its ever
becoming a science’.6
Sartori (1984, 59) notes that Popper’s proposition that ‘precision’ may be a false
ideal was actually a remark on the precision of mathematical reasoning, but had
nothing to do with commending imprecision for natural languages. Sartori says, ‘it
is a wholesale distortion to call upon Popper to support a crescendo of termino-
logical and logical sloppiness’. The idea is not that language can be frozen, but the
specialised language we need for social science discussion needs to be relatively
stable.
The discussion of categories and labels in this article sticks to a rather pedestrian
dimension that underlines the difficulty of counting and interpreting without
agreed metrics, but clearly the problems are reinforced if one pursues discussions
of semantics in the rather more ambitious fashion of Sartori (1984). Sartori (1984,
16) notes that what is not named largely remains unnoticed and, moreover, that the
naming choice involves a far reaching interpretive projection. One can crudely say he
is drawing attention to the consequence of choices of categories and labels. In
Sartori’s terms (1984, 29) we need disambiguation. This article is predicated on the
assumption that if we cannot do some micro scale under-the-stairs ‘terminological
house-clearing’ (Sartori 1984, 62) vis à vis a simple concept such as interest group,
we are unlikely to manage the rambling mansion of the fuller discipline. The
assumption is that particular aspects of the subject need to be resolved before real-
istically attacking a broader front.
198 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
Sartori discusses the work of Whorf (1956) who argues that in part language influ-
ences (not determines) thinking: in other words thoughts tend to accept linguistic
‘givens’. As Sartori (1984, 18) puts it, natural language is ‘thought moulding’ in a
way that vocabulary embodies perceptions and conceptions of reality. While most
words are ambiguous, the context can usually clarify meaning, but that is not the
case with the terms in this article. He considers semantic terms to have two pos-
sible aspects—slicing and interpretative. In political science there is a sort of slicing
vocabulary that permits the world to be parcelled up in preliminary fashion. There
would be communication difficulties without rough and ready acceptance of a
common language of political party, interest group, parliament, etc.
But of course this common language is deceptive in three ways. In the simple way
that concerns this article, the boundaries are simply not agreed if one tries to pursue
specialist, detailed discussion. Secondly, Sartori notes that the selection of ‘slices’
is likely to reflect potential use. Thus the famous example of the Eskimos (sic) who
did not have a single word for snow, but a dictionary of terms for the specific states
of snow, while the Aztecs had one term to cover ice, cold and snow. Thus the argu-
ment is that the slices are not reflections of nature but cultural selections. Areas of
specialism need a specialist terminology not as jargon or to make matters arcane,
but simply because relevant distinctions cannot be captured with categories that
are too blunt.
If one is convinced by Sartori then there is a problem of missing literatures. There
simply is no discussion along the lines he suggests in the interest group field. On
the other hand there is a danger of what he refers to elsewhere (1970, 1033) as
logical perfectionism (and consequent paralysis of research). The discussion of
terms cannot be the end of the exercise. The clarification of terms is to permit
empirical work rather than replace it.
The Consequence of Conflicting Images of
Interest Groups
Definitional issues are central to securing efficient reward from the collective effort
of the research community. Two very broad styles of interest group definition have
evolved. One essentially focuses on the normative (social capital) benefits of indi-
vidualised voluntary participation. The other is described below as functionally
driven.
(i) The Traditional Voluntary Interpretation
Researchers in the voluntary tradition would probably present groups as voluntary,
democratically accountable and individual-based. A definition compatible with this
interpretation would include most of the following features. Interest (voluntary)
group stereotype:
(i) it is organised only for a specific collective political end (such as the aboli-
tion of slavery);
(ii) its goal is attainable; the group can be disbanded on realisation of the goal;
DEFINING INTERESTS 199
(iii) it is a non-governmental body;
(iv) it does not seek to form a government itself, but to influence public policy;
(v) it has a formal, mostly voluntary, membership;7
(vi) the membership has some control over the leadership of the organisation
in connection with goals and means (internal democracy);
(vii) the membership funds the organisation through subscriptions;
(viii) it is organised to give expression to shared attitudes or pursue shared inter-
ests (promotional/sectional) of a given constituency;
(ix) as membership is seen to reflect a matching of group and individual goals,
then it can be expected that members join groups with some long term
attachment.
Any body with (or assumed to have) most of the sorts of qualities listed above
would be defined as a pressure/interest group by almost any source. When asked
for examples of interest groups most students and others cite Oxfam, trades unions,
Greenpeace—individual cases that (roughly) fit the ‘voluntary’ interpretation.
Moreover it is membership-based voluntary organisations (or to be more pedan-
tic, those that are assumed to have members)8that are at the heart of the central
debate in the group literature. How do groups address the ‘logic of collective action’
issue? How do such groups encourage participation in groups unless by selective
incentives?
These characteristics also reflect many expectations about the political system and
its reliance on groups: groups are seen as replacing political parties as vehicles of
democratic participation. If the wider population of groups in practice lack most of
the characteristics of these examples, then commonly held views on the legitimate
basis of group participation in policy-making and governance are undermined. If
group influence in policy-making is to be normatively acceptable, then they are
probably expected to demonstrate the ‘representativeness’ and internal democracy
at the heart of the above portrait (Grant 2001, 348).
However, many other bodies regarded as groups in important sources in the liter-
ature, predominantly in the United States, lack many of these features. (Below these
sources are discussed as being ‘policy studies centred’.) For example: (i) for some
important interest groups political activity may not be their primary activity; (ii)
civil servants, government departments, firms, etc. can be perceived as having a
policy interest and influence; (iii) some groups take little trouble to recruit
members; (iv) internal democracy may be largely symbolic;9(v) a group may be
funded by its supporters or patrons rather than its formal membership; and (vi)
groups may be collective structures for ‘organisations of organisations’ rather than
individuals. Many important groups are collections of institutions—companies,
local authorities or other non-individuals (Salisbury 1984).
Graham Wilson’s initial interpretation of an interest group reflects the ‘stereotype’
version above. He noted (1990, 6),
some organisations are clearly what we should understand an interest
group to be. Such organisations recruit members with the explicit promise
that the organisation will pursue certain public policy goals. People who
join the League Against Cruel Sports in Britain or Common Cause in the
200 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
United States are clearly joining together to support an organisation
which is obviously dedicated to certain relatively well-defined public
policy objectives.
While Wilson’s examples match the above stereotype, he warns, ‘Few organisa-
tions commonly regarded as interest groups are like the League or Common Cause’.
In line with Wilson this article asserts that the organisations that are most significant
in particular policy processes tend not to conform to the stereotype that best approximates a
pressure/interest group. The paradox is that the most influential organisations are in fact
highly likely to be the bodies furthest from the pressure group stereotype set out above. In
other words, studies by David Lowery and Virginia Gray (1995) and by other
authors show (in a language suggesting interest group activity) the dominance
within the political process by corporations and business ‘organisations of organi-
sations’ rather than individual-based groups that the term ‘interest group’ popu-
larly suggests. The most influential types of interest groups, as defined in public
policy studies, are most distant from the core of the image of interest groups as
deployed by many dedicated interest group researchers. It is this divergence in
expectations amongst interest group and public policy researchers that underscores
the importance of agreeing what is studied when interest groups are studied. It
may be that ‘conceptual stretching’ (Sartori 1970, 1034) is inevitable, but the
stretched term is not synonymous with earlier formulations.
(ii) The Policy Studies Use: Pressure Group by Function?
In line with the previous section, Heinz et al. (1993, 29) note that, ‘for years schol-
ars presumed that interest groups meant voluntary associations, Truman’s (1951)
magisterial work deals entirely with such organizations, and they are the focus of
Mancur Olson (1965) and those who have followed his lead. Most of the abun-
dant case studies of interest group activity also examine groups whose members
choose, more or less freely, to support the group’. However Heinz et al. note the
transformation of the field of inquiry: ‘subsequently, the analysis of interested
activity was explicitly extended to include organizations that had long been active in seeking
to advance their interests ... business corporations, universities, state and local gov-
ernments’ (emphasis added). This extension of the scope of coverage in the policy
field is justifiable in a policy studies context, but it reads oddly for those who regard
the central interest group issue as being about mobilising individuals for collective
action (the Olson paradox).
The policy studies literature tends to accept as a pressure/interest group any organ-
isation that is seen as being active in the policy process with the function or aim of
influencing policy outcomes. The conclusion implied by this line of argument is that a
pressure group may be defined by function: any body that seeks to influence policy
is a pressure group. The most important consequence of this approach is the fact that in
this light an individual company can be seen as an interest group. Yet companies, as will
be outlined below, are different organisational animals. They do not have members,
which means they experience none of the collective action issues foreshadowed
by Olson, and they lack the potential to imbue the political process with higher
levels of citizen participation in decision-making.
DEFINING INTERESTS 201
This functional sort of definition extends the term interest group from the core
coverage of groups like the British Medical Association to include not only com-
panies but also bodies such as a government ministry or department, public cor-
poration or local authority. Robert Salisbury (1984, 64) showed how organisations
of the stereotypical ‘pure’ style might be numerically few and politically margin-
alised in policy processes. He noted the policy-making importance of state and local
governments, corporations, think tanks and lobbyists; for some authors these are
now part of the interest group terrain. The public policy perspective of policy-active
organisations has slipped across into general definitions of interest group.
Wilson (1990, 7) asked, ‘are we to conclude that any organisation which seeks to
any degree to influence public policy is therefore to be regarded as an interest
group?’ His (and for the policy studies sources generally) conclusion was a broad
‘yes’. He noted that many studies of interest groups defined the term narrowly,
and have focused only on membership organisations: ‘such a narrow definition of
interest groups would have the peculiar effect of excluding large businesses such
as ICI or Du Pont from our view, surely a strange and unwelcome development’.
For Wilson (1990, 8) an interest group is an organisation which has some auton-
omy from government or political parties and that tries to influence public policy.
This is conspicuously distant from the ‘voluntary’ stereotype outlined above.
Charles Lindblom adopted a similarly wide-brush approach (1980, 84–85):
the study of interest groups now also identifies government officials, their
associations, and their departments or agencies as playing interest-group
roles.
This led Lindblom (1980, 85) to ‘stake out’ the broadest use of the term in that:
we mean by interest-group activities all interactions through which
individuals and private groups not holding government authority seek to
influence policy, together with those policy-influencing interactions of
government officials that go well beyond the direct use of their authority.
The consequences for comparability of some researchers working with such a dif-
ferent tool are obvious. Jeff Berry’s (1989, 191) work confirms that many key
policy-relevant organisation are not interest groups but corporations. In practice
Berry (1989, 20) accepted a broad use of the term interest group. He saw it as
embracing ‘corporate offices, trade associations, farm groups, labor unions, politi-
cal action committees, law firms and public relations firms’. Similarly, data used
by Gray and Lowery (1992) on measuring the diversity of interest groups was based
on the number of groups (sic) ‘registered to lobby in the state legislatures’. This
data was used because of its reliability and accessibility but the lists used ‘include
membership organization (e.g. unions, professional associations, civil rights
groups), corporations, trade associations, churches, and governmental representa-
tives. Thus, organized interests is a more appropriate term than interest groups since insti-
tutions and non membership organizations are counted as well as voluntary associations’
(emphasis added) (see similar use in Lowery and Gray 1995). While this reserva-
tion is made clear by Gray and Lowery, nonetheless the findings percolate into the
literature as statements about interest groups. Likewise the availability of data from
202 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
lobbying registration appears to have driven as a matter of scholarly convenience
a particularly broad group interpretation in the United States.
There are however collective costs to the discipline in pursuing an ad hoc pattern
of labelling regardless of how well the terms are defined by individual researchers.
Firstly, there is an obvious tediousness if we must relearn terms anew in every
context: the purpose of a professional language is precisely to ease communication
by building up a stock of transferable terms. Secondly, the possibility of compari-
son is seriously affected if non-standard definitions are adopted. Thirdly, charac-
terising an interest group on the basis of an observable policy-related function
obscures key differences in the internal dimension of activity. That is, it inhibits
researchers examining whether the (internal) organisational and collective action
issues of the group are similar or different, existent or non-existent.
It should be obvious that the thrust of this argument is not that corporations are
not relevant to the policy process. It is to say they are not relevant as interest groups.
If a student essay answered a question on interest groups by discussing, for
example, Shell, foreign currency markets and British Airways, would there not be
a rather significant ‘hole’—the collective action issue? This article tries to give us
the facility to distinguish the issues raised in the complex interaction of institutions
in modern policy-making with the field of collective action issues signalled by Olson
and others. The first are interesting problems that must be addressed in political
science, but why call them interest group problems?
Groups as Membership Bodies?10
The central tension here is that there are two interest group literatures. That dealing
with membership and intra-group issues tends to use as a starting point the vol-
untary stereotype. ‘Interest group-ness’ in such sources sees the mobilisation of
members as the key issue. Such a position is different to the policy studies view
that primarily sees aspirations to policy influence as the key requirement for inter-
est group status. Salisbury (1992, 41) notes that ‘a corporation, a local government,
most churches, and even universities are different, not totally but in crucial ways,
from our conventional notion of interest groups’ (emphasis added).11 Among the
important differences he records is the fact that interest group members have the
possibility of ‘exit’ as a means of influencing organisational leaders. He says, ‘insti-
tutions are managed organizations ... membership groups must look far more care-
fully to the desires of their members, both of assure political legitimacy and to keep
their supporters happy’ (ibid., 43). If, as Salisbury claims (ibid., 36), there are
important empirical differences between a system driven by membership groups
and one in which institutions occupy centre stage then there must be adverse con-
sequences in uses of the term interest group that run together.12
In dealing with the issue of interest group numbers and growth, Baumgartner and
Walker (1988, 908) note, ‘the first step in discovering how many members the
American group system actually has must be to insure that our sources of infor-
mation are reliable’. Their study is of ‘membership of voluntary associations’. The
definitions used and the scope of the study are clearly set out and again the issue
is not about weakness in its own terms, but to note that their discussion of ‘the
DEFINING INTERESTS 203
changing shape of the interest group system’ is very different from the Lowery and
Gray (1995) terrain because the scope of the interest group term is very different.13
Their sort of definitional assumption about the multi-member organisation seeking
collective ends ties in with the debate on the logic of membership (Olson, 1965).
Pressure/interest groups in such an interpretation are only those bodies which
attempt to influence policy and are multi-member organisations seeking collective ends.
This ‘seam’ of work, in contradistinction to the policy studies approach, sees the
membership issue as crucial. Jack Walker (1991, 4) noted,
my principal focus is limited to functioning associations in the United
States that are open to membership and are concerned with some aspects
of public policy at the national level.
Walker is also clearly a ‘group as collective phenomenon’ source as opposed to a
‘group by function’ author. He noted that,
by focusing on groups that admit members, I excluded the public affairs
divisions of private corporations such as Ford Motor Company, Sears,
Roebuck and Co, General Electric, or Xerox.
This exclusion, of course, is precisely the sort of policy-influencing body that
Schlozman and Tierney, and others in the field with a policy studies orientation,
included. Kay Schlozman (1992, 818), in her review of Walker (1991), makes the
point that ‘Walker has no data about the thousands of organizations active in Wash-
ington (e.g. corporations, universities and think tanks) that have no members in
the ordinary sense’. The consequence of the definitional scope adopted by Walker
is that while his work is still of compelling importance, one cannot generalize from
the pattern of interest groups he discusses to make arguments about the weight of
business pressure in the political process. His map (deliberately) leaves out those
many policy-relevant companies acting on their own account. This is a strength if
one’s focus is membership organisations, but a weakness from a public policy per-
spective concerned with explaining influences on policy outcomes.
Groups: Individual-Based?
Even if the ‘interest group’ label were restricted to membership bodies, some would
want to make further restrictions to it being used to apply to ‘organisations of indi-
viduals’ (in line with the stereotype described at the outset of this article) rather
than ‘organisations of organisations’. Joseph LaPalombara (1964, 18) defined an
interest group as ‘any aggregation of interacting individuals who manifest con-
scious desires concerning the authoritative allocation of values’. Berry’s (1984, 5)
definition of an interest group is ‘an organized body of individuals who share some
goals and who try to influence public policy’ (emphasis added).
These sorts of definitions exclude the ‘organisations of organisations’ such as busi-
ness groups and thus connect with the bulk of the literature about the rationality
of membership which on the whole concerns membership decisions by individu-
als. But though the issues of membership are different for organisations of indi-
viduals and collectivities of other institutions, the mainstream use of the term does
204 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
not confine itself to looking at organisations of individuals. In practice, Berry
himself accepted that trade associations and other bodies were interest groups. That
his formal definition is narrower shows the power of the Olson paradigm about
individuals and collective action in shaping the way we look at groups. Berry’s
practice is more commonly followed than his definition. Such an extension of the
membership issue from individuals to all organisations seems irresistible, because
as a broad generalisation it appears that the most powerful groups in the policy
process are not based on individual members. The stereotype is again defective. The
influence of private sector firms and their organisations is probably greater than
that of consumers and their groups. Studies of politics would not want to ignore
the former.
The membership issue for companies is, however, different than for individuals. One
cannot assume that discussions of the membership of companies in trade bodies
(numerically a large percentage of the bodies conventionally regarded as interest
groups) are analogous to the decisions made by individuals about membership of
cause or sectional groups. But nonetheless a definition restricted to individuals is
too narrow.
Must Groups be Organised?
Clive Thomas and Ron Hrebenar (1990, 124) defined an interest group as ‘any
association of individuals, whether formally organized or not, which attempts to influ-
ence public policy’ (emphasis added).14 At first sight this coincides with the Berry
definition (above) with its stress on individual membership, but in fact there is an
important difference which again underlines the fact that care is needed in reading
across the literature. They include in their interpretation groups of individuals that
lack formal organisation. Many other sources accept the ‘either organised or not’
character including Geoffrey Alderman (1984, 21). Thomas and Hrebenar are self-
consciously explicit in their work. The issue is that even if everyone else was as
clear in their separate and defensible definitions (and they are not) the overall lit-
erature would be unsatisfactory. If there is not a common currency of terms then
it is difficult—and even dangerous—to generalise across the literature. Different
sources are discussing different phenomena. If unorganised associations of indi-
viduals are ‘interest groups’, then the idea of some kind of census or map is made
more difficult, if not impossible. There is thus no deep agreement on whether
groups must be organised to fit the field of study—or what groups are considered
to be part of the category of organised interests. The notion of the non-organised
group spreads the definitional ambiguity into the zone of social movements.
New Terms, New Focus? Distinguishing Policy
Participants and Pressure/Interest Groups
This part of the article reinforces that there are very different interpretations sub-
sumed by the interest group label (and therefore real difficulties in addressing the
literature as a whole as there are so many definitional uses), and then offers a new
‘slicing’ in Sartori’s terms. As signalled earlier, this section is required only if one
DEFINING INTERESTS 205
assumes the existing repertoire of terms is imperfect. And the picture of imperfec-
tion can remain valid even if this redrawing of zones is unconvincing.
As we have seen above, the broad functional approach assumes that if an organisa-
tion (even a corporation) is involved in the policy process, it is (de facto) a pressure/
interest group. In practice the ‘de facto’ qualification becomes ignored. Richard Rose
(1985, 247, see also Baggott 1995, 11) resists this slippage from interests to interest
groups and complains that because the word ‘group’ is a very general term, writings
about pressure groups sometimes treat every political institution as a manifestation
of pressure politics. He argues that the ‘definition by function’ is a reductio ad absur-
dum because it blurs all distinctions between those responsible for exercising the
powers of government and those groups that seek to influence government but not
hold public office. However it is precisely the thrust of those who pursue the func-
tional argument that the formal boundaries are misleading and that parts of the
administration are in group-like competition with other sections of the bureaucracy.
Often the groups outside will share aims and tactics with ‘patrons’ in the civil
service.
The argument in this article is that the real confusion is not caused by assuming
that almost any kind of organisation will attempt to exert influence, but in labelling
every organisation engaging in pressure politics as a pressure group. If all bodies
attempting to affect outcomes in the ‘pressure system’ are given a generic name
such as pressure participant, there is no need to call every policy-influencing body
an interest/pressure group. There can be a broad equivalence in functional policy
terms between different types of policy-influencing bodies without them all being
lumped together. In Sartori’s language (1984, 30) this is a denotative definition: it
establishes the broad boundary of the class.
This article acknowledges that there are many bodies active on most policy fronts
but, like Rose’s work, resists (in spite of Jordan and Richardson 1987) the idea that
all these bodies are thereby interest groups. While the list of policy-relevant active
organisations can be captured by a term such as pressure participant, within that
category a distinction can be made between policy participants such as companies
and a more restricted conception of interest or pressure group. This new term (policy
participant) is needed to signal that not all bodies seeking to influence policy are
sensibly lumped into the pressure group box. Attempting to influence public policy
as a pressure participant does not make all such bodies ‘pressure or interest groups’.
This in the Sartori scheme is a ‘precising definition’.
It is less confusing to see the use of the term ‘pressure participant’ for the wider
conception and reserve the terms ‘pressure/interest group’ for (normally) multi-
member, politically oriented bodies of individuals (in this case both citizens and com-
panies as business group members), and ‘policy participant’ for bodies such as a
company, lobbyist or agency active in the political system. If ‘interest group’ is
allowed to take the wider meaning, it causes confusion for students who are
directed to the Olson-type literature on the internal life of groups: such issues
do not arise for all pressure participants. This ‘rule’ allows a distinction between
the (collective) trade association with companies in membership (such as the
Confederation of British Industry or the Federation of Small Businesses) from
own-account lobbying by individual companies (whether through hired lobbyists
206 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
or in-house public affairs teams). Both might attempt to influence relevant legis-
lation, but the single company does not have the membership/policy problems that
are at the heart of mobilisation-centred interest group studies. Companies in their
policy-influencing role are pressure participants in that they are active in the policy
process, but they can be categorised as policy participants rather than considering
them as interest groups simply on the basis they are seeking to alter policy.
From a policy studies perspective it is important to examine firms both in the (col-
lective) pressure groups that they join and as ‘own-account’ actors. However, many
interest group studies have as their primary goal investigation of the collective action
paradox. For these studies the focus will not be on organisations acting politically
on their own account but acting, for example, through the trade association or the
Eurogroup. A study of all policy-influencing bodies would not require to distin-
guish between the firm and the collective business pressure group. Both would be
pressure participants.15
To sum up again, the policy studies literature effectively homogenises all ‘political
organisations’ who seek policy influence, and appropriates for them the label pres-
sure/interest group. This is reflected in the label pressure participant—which can
be subdivided into collective action interest groups and policy participants. This
confirms what interest group scholars have long acknowledged, which is that all
groups are not the same and that underlying their policy function is a heteroge-
neous array of organisational and representative functions. A language is needed
to capture the distinctions.
Memberless Groups: an Exception to
the General Argument?
The main thrust of the interest/pressure group definition advocated here is that it
is a multi-member body attempting to influence policy. However a second sort of
interest group making explicit policy-influencing efforts can be identified—the
policy-centred group. Such bodies have few members (and hence no real politics of
membership), and exist only to influence policy and would seem to meet most com-
monsense interpretations of ‘pressure group’. Even though such organisations do
not face the politics of collective policy-making associated with ‘pressure groups’,
it would be strange to exclude this kind of body from the pressure group category
as pressure is its raison d’être. In line with this proposition Michael Hayes (1986,
133) argues that,
interest groups are typically portrayed as organizations that seek to mobi-
lize members to influence public policy. Numerous groups formed in
recent years, however, do not conform to this traditional conception of
the mass-membership group. Many have no members at all ... some orga-
nizations are composed of nothing but staffers in Washington and are
funded entirely by sources such as private foundations.
Accepting Salisbury’s argument that interest groups are distinctive because
members (or supporters for that matter) can potentially exercise control of leaders
through ‘exit’, the memberless group is a sort of affront to that logic, but the policy-
DEFINING INTERESTS 207
centred organisation—with its main priority of influencing public policy—does
seem worth including under the interest group banner. Thus organisations like the
Campaign for Lead Free Air (CLEAR) in Britain, who campaign for the removal of
lead from petrol (see Wilson 1983), are included as interest groups even though
they are not a membership organisation. Essentially the policy-centred group is a politi-
cal (non-collective) organised group. It is an organisation set up with the express purpose
of influencing policy.
Pressure Participant, Interest Group or
Policy Participant Status
The underlying image of the political process adopted here is one where policies
emerge from the interaction of parties, bureaucrats, companies, membership-based
groups, trade bodies, groups with few members, competing elites and public
opinion. This view was captured by Lindblom (1979, 523) when he presented poli-
cies as being the result of mutual adjustments. This interaction may be described
as between pressure participants (or stakeholders) rather than between widely
defined (functional) pressure or interest groups. Pressure participants are bodies
attempting to influence outcomes. Thus a university is normally regarded as ‘politi-
cal’ only in the very broadest sense when it is engaged in the delivery of core edu-
cation; it is however a pressure participant when, for example, it intervenes on its
own account in the political process if there is a cut in the level of resource allo-
cated. In line with the argument above, the university can at that point be termed
a policy participant—an apolitical body that sometimes acts in ways that from a
functional perspective make it appear as if it is a pressure group. That political
science treats organisations such as the Church of Scotland as ‘political pressure
groups’ simply proves that there is nothing better to hand. Similarly, once the new
terminology is adopted, a business corporation (e.g. British Petroleum) which has
its own government affairs section need not be seen as a pressure group—even
when it seeks to intervene politically—but it is undoubtedly a pressure participant
(and more narrowly policy participant). In the UK, the Caravan Club is primarily
an organisation to provide members with access to quality camping sites and facil-
ities. Were it to act politically, say to lobby for planning changes to open up new
areas for camping facilities, it would be a pressure participant. To distinguish
the politically active university, a company such as BP or the Caravan Club from
other pressure participants (of the interest group type) they can be labelled policy
participants.
Potentially an already organised body with no prior political orientation could
become politically involved. That is, it could shift from being a potential pressure
participant to a pressure participant. Whether it is an organised, collective group
(a club or society) or an organised, non-collective body (e.g. firm) it is a policy par-
ticipant. These are discussed by Robert Dowse and John Hughes (1977) as ‘spo-
radic interventionists’. An example of this type is the National Coursing Club (NCC)
in Britain. This collective body had minimal political relevance until the contro-
versy of the hunting debate refocused public attention on the practice of hare
coursing. The term policy participant allows discussion of such organisations as rel-
evant to the policy process—without having to grant them ‘honorary’ and tempo-
208 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
rary status as interest groups. They are not interest groups, but adopting a role
acting as if they are. Such interventions can be politically very significant, but not
interest group phenomena.
The Countryside Alliance—created for the sole (political) purpose of influencing
related legislation—is a pressure group, while the NCC are acting as if they are a
pressure group. The NCC’s political activities are additional to its other activities
and are likely to be ephemeral (i.e. they will not become a permanent feature). It
is possible to label the clubs as temporary pressure groups. Such groups are, more
precisely, intermittent/occasional policy participants. But this is a difficult bound-
ary: some would accept the collective body which is temporarily politically active
as a pressure group—even though that is not its primary purpose. In the terms of
this article though it is regarded as a policy participant. It is a virtue of these terms
that one does not have to use the label ‘pressure groups’ to cover both collective
groups such as the Friends of the Earth and a local amenity club forced to act defen-
sively, and temporarily, in the realm of politics on a particular issue.
Policy histories confirm that business organisations are particularly important politi-
cal actors. They are often politically relevant—both when acting through trade
associations (which would make them subject to conventional pressure/interest
group interpretations) and when acting on their own account (e.g. through gov-
ernment affairs departments). As argued in this article they are policy participants
when acting on their own account. The major point is that the range of influence-
seeking organisations that would be of relevance in a study of policy outcomes is
different from the range of bodies that are usually (but not always) considered as
pressure/interest groups. The latter are (primarily) characterised by the need to
overcome the free rider problem and to settle on collective policy goals.
Implications for Research: Working with
these Categories
The literature is littered with idiosyncratic deployments of the term interest group.
The article prefers to reserve this term for dedicated politically relevant organisa-
tions (normally with members or supporters) and use policy participant for organ-
isations not primarily dedicated to political activity. The sum of these two categories
is the pressure participant category. All pressure participants are equivalents when
captured by the functional view dominant in public policy studies.
So what is the benefit of these new categories? They remind us about the diver-
sity within the pressure participant category as documented in public policy studies.
Deployed as suggested we may be able to retrieve, from what has been up till now
referred to as ‘interest group literature’, those conclusions that are relevant to what
we refer to as interest groups and what are relevant to policy participants. This
organising language can potentially allow us to squeeze more from our existing
research efforts. As such they may assist in addressing the criticism, made by some
prominent interest group scholars, that inconsistency in research frameworks has
thwarted attempts to generate solid and enduring conclusions about interest group
activity. Perhaps most importantly, they remind us that questions about interest
groups, such as the asserted dominance of business, should not be answered by
DEFINING INTERESTS 209
data on pressure participants. Similarly, the categories remind us that only a minor-
ity of what have been in the past termed ‘interest groups’ are actually in a posi-
tion to even potentially enhance political participation. The contribution to direct
citizen participation in democracy is not enhanced through policy participants.
Likewise, the social capital development capacity may be small when policy par-
ticipants are withdrawn from the ‘interest group’ category.
The use of the labels advocated does imply additional work for researchers. This is
primarily because the names of groups do not immediately assist us in the cate-
gorisation task. A particular danger, with a capacity to corrupt data gathering, is
the fact that some organisations have retained the language of mutuality when
they have de facto become service industry firms—for example roadside assistance
motoring organisations. Other commercial organisations use terms such as mem-
bership but the ‘membership’ confers nothing other than a right to use facilities—
e.g. gymnasia. The additional work involved in assessing how groups fit into these
categories, combined with the proliferation of policy participants, makes for a large-
scale task, but surely one that is routine for researchers trying to come to grips
with their subject. This type of classificatory task is a necessary prerequisite for
deriving conclusions that carry beyond the case.
Summary
Confusion has been caused by the policy studies practice of labelling everything
that attempts to influence policy as a pressure or interest group. The practice arose
for good reason: there is a common ground between all policy-influencing organ-
isations in a functional sense. However, it has had the repercussion of treating all
policy participants as though they had the same levels of flexibility, decision-
making processes informing their policy behaviour, autonomy and resource gen-
erating capacity as one another. This article argues that common ground can be
recognised by terming all these bodies pressure participants (a term that picks up
some of the early use in the literature). Within that general category the unitary
interest policy participant is distinct from the collective interest or pressure group. The
differences between the phenomena are too great to sensibly combine. The costs
of failing to make distinctions are too great to ignore. The new approach offers the
opportunity for worthwhile comparative work that is denied if ad hoc definitions
generate data sets with no portability.
John Gerring (1999, 357) argues that ‘good’ concepts are not synonymous with
clarity or the product of a particular formative process, but that they are a trade-
off among many desirable dimensions (familiarity, resonance, parsimony, coher-
ence, differentiation, depth, theoretical utility and field utility). Whether the terms
introduced in this article meet this kind of test can be debated: that the terms used
unthinkingly fail these tests is beyond dispute.
New terminology is introduced with reluctance. As Felix Oppenheim argued (1975,
307) there are advantages in the specialist discussion of politics remaining reason-
ably close to ordinary language, but he went on to quote J. L. Austin’s (1970, 185)
advice, ‘ordinary language is not the last word; in principle it can everywhere be
supplemented and improved upon and superseded. Only remember, it is the first
210 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
word’ (original emphasis). Gerring (1999, 362) notes deviations from ordinary lan-
guage impose costs, but they are necessary departures. The coincidence that the
ideas required are those inherited from general language is too unlikely. It would
be remarkable if a general conversational use of the term pressure group satisfied
the needs of political science. The distinction between pressure/interest group,
pressure participant and policy participant might then be useful.
There may be criticism that to focus on definitional clarity amounts to ‘pickiness’.
We disagree. The motivation is that the profession is currently making do with ill-
defined concepts that hinder the extraction of gains from our collective research
efforts. Indeed, contra to objections that this is ‘too much’, more realistically this
should be viewed as preliminary to a more thoroughgoing attempt to develop more
robust and theoretically informed categories. The valid criticism of this article is
not that it overly complicates analysis, but that it is too modest a start on what
should be a bigger reconceptualisation of the tools of the sub-discipline.
Richardson (1999) asked why British work on groups and parties had so little
impact and did the ‘haze of common knowledge’ obscure? This piece suggests it
does. If we have no international language then our studies of groups in Britain
are in effect unintelligible outside their locality. He quotes Mackenzie’s call for the-
ories and concepts that can be utilised in different national contexts. Yet again
Mackenzie beat our generation to the punch. The sort of impressions we have
about the relative power of interest groups are dependent on the way in which
‘interest group’ is defined and how it is operationalised. Noting the bias to middle-
class participation in groups, Schlozman (1984) has picked up Schattschneider’s
well known observation about the interest group system being congenitally biased
in a middle-class way and asked, ‘what accent the heavenly chorus?’. At least in
part the answer depends on who is conducting the research and their conceptual
interpretations—because that determines who is counted as singing.
About the Authors
Grant Jordan, Department of Politics & International Relations, University of Aberdeen,
Scotland, email: g.jordan@abdn.ac.uk
Darren Halpin, Centre for Social Science Research, Central Queensland University, North
Rockhampton, Queensland QLD 4702, Australia, email: d.halpin@cqu.edu.au
Dr. William Maloney, Department of Politics & International Relations, University of Aberdeen,
Scotland, email: w.a.maloney@abdn.ac.uk
Notes
1. For example, governance, corporatism, globalisation.
2. This is based on a British Interest Group Project Working Paper, G. Jordan, W. Maloney and
A. McLaughlin, ‘What is Studied When Pressure Groups are Studied’ (1992).
3. As Salisbury (1975, 176) puts it, ‘Those scholars who continue to employ “pressure group” as their
generic term mean by it the same as those who prefer the emotively more neutral term “interest
group” ’.
4. The distinction between interest groups and New Social Movements is discussed in Jordan and
Maloney 1997, chapter 2.
5. A range of labels are used almost interchangeably to refer to NGOs which makes an authoritative
definition hard to find. These labels include Nonprofit, Voluntary, Social Economy, Third Sector
DEFINING INTERESTS 211
Organisations, Community Organisations, Community Sector Organisations and Civil Society Organ-
isations (Lyons 2001, 8–9).
6. Sartori actually suggests Wittgenstein is correct in saying that meaning and context are interwoven,
but there is a different question of how meanings are treated in the specialised vocabularies (lan-
guages) that each science constructs. Sartori also notes the extension of the Wittgenstein fallacy to
say that meaning is revealed by context only, ‘the more meanings that must be surmised by con-
textual investigation, the less we are dealing with a scientific kind of knowledge’.
7. The distinction between individuals and businesses is blurred when it comes to organisations repre-
senting small businesses. The businesses may be individuals.
8. Important campaigning examples may well not have the stereotypical arrangement of formal mem-
bership and rights within an organisation. Thus it would be unfortunate if our conception of cam-
paigning group did not include Greenpeace—which lacks formal members.
9. The Royal Society for the Protection of Birds, for example, argues that they make policy not by
responding to their postbag but through ‘proper, structured, unsolicited surveying, which allows us
to make the right decisions’ (Daily Telegraph, 23 April 1994).
10. Salisbury (1992, 43) also sees membership as a crucial dimension.
11. Salisbury (1992, 44) also refers to ‘conventionally defined interest groups’. Arguably the stereotypi-
cal definition given in this article is near to what he had in mind.
12. Mitchell (1990) uses the corporation rather than the interest group as his object of study and there-
fore is able to pursue Salisbury’s distinction between the institutional and group representation. In
fact in real life there are more ‘membership’ groups than membership groups. Many groups that
look superficially to be member-based are in fact supporter-based with regular financial contributors
rather than members with rights within the organisation.
13. Salisbury (1992, 57) says, ‘Business corporations are sometimes considered, sensibly enough, among
business interests, but the profound differences in organizational structure and motivation between,
for example, General Motors and the Chamber of Commerce have not been remarked, nor has the
implication these differences have for interest group theory’. The position in this article is that in
policy studies terms the differences may be comparatively unimportant, but if one is interested in
the mobilisation issue that has dominated interest group studies since Olson (1965), then the dif-
ferences are vital and the corporation cannot be usefully subsumed within the interest group
umbrella.
14. See also, ‘those units, organised or not, of the democratic process which have a set purpose or set of
purposes, but which are nonetheless neither political parties nor formal agencies of governmen’
(Alderman 1984, 21, cited in Baggott 1995, 3, emphasis added). Grant (1995, 6) suggests that ‘Jordan
and Richardson (1987, 14–18) are correct when they argue for a broad definition of pressure group
that accepts companies and corporations as such groups’. In the passage alluded to by Grant, Jordan
and Richardson cited the miners’ strike of 1984–1985 to show that in policy studies sense it is easy
to subconsciously, and defensible to consciously, stretch the group concept. See also discussion by
Salisbury (1984, 64). This article obviously does not follow the Jordan and Richardson policy-based
view.
15. Cigler and Loomis (1991, 385) note that ‘because non group, special interest actors increasingly enter
the fray as well, the term organized interests may well be more applicable and make more sense today
than the term interest groups’. In other words, the line of argument of this article is already in the lit-
erature in a low-key way.
Bibliography
Alderman, G. (1984) Pressure Groups and Government in Great Britain (London: Longman).
Baggott, R. (1995) Pressure Groups Today (Manchester: Manchester University Press).
Baumgartner, F. R. and Leech, B. (1998) Basic Interests: The Importance of Groups in Politics and Political Science
(Princeton, NJ: Princeton University Press).
Baumgartner, F. R. and Walker, J. (1988) ‘Survey Research and Membership in Voluntary Organizations’,
American Journal of Political Science, 32: 908–928.
Berry, J. M. (1984 and 1989) The Interest Group Society (Boston: Little, Brown and Company).
Cigler, A. J. and Loomis, B. (1991 edn) Interest Group Politics (Washington: CQ Press).
Dowse, R. and Hughes, J. (1977) ‘Sporadic interventionists’, Political Studies, 25:1, 84–92.
Gerring, J. (1999) ‘What makes a concept good?’, Polity, XXI:3, 357–393.
212 GRANT JORDAN, DARREN HALPIN, WILLIAM MALONEY
Grant, W. (1989, 1995) Pressure Groups, Politics and Democracy in Britain (Hemel Hempstead: Philip Allen).
Grant, W. (2001) ‘Pressure politics: from “insider” politics to direct action?’, Parliamentary Affairs, 54:2,
337–348.
Gray, V. and Lowery, D. (1992) State interest group system diversity’, Political Research Quarterly, 46:1,
81–97.
Hayes, M. T. (1986) ‘The new group universe’, in A. J. Cigler and B. Loomis, Interest Group Politics
(Washington: CQ Press).
Heinz, J. P., Lauman, E. O., Nelson, R. L. and Salisbury, R. H. (1993) The Hollow Core: Private Interests in
National Policy Making (Cambridge: Harvard).
Jordan, G. and Maloney, W. A. (1997) The Protest Business? Mobilising Campaign Groups (Manchester:
Manchester University Press).
Jordan, A. G. and Richardson, J. J. (1987) Government and Pressure Groups in Britain (London: Oxford
University Press).
Kimber, R. and Richardson J. J. (1973) Pressure Groups in Britain (London: Dent and Sons).
LaPalombara, J. (1964) Interest Groups in Italian Politics (Princeton, New Jersey: Princeton University Press).
Lindblom, C. E. (1979) ‘Still muddling: not yet through’, Public Administration Review, Nov/Dec, 517–526.
Lindblom, C. E. (1980) The Policy Making Process (Englewood Cliffs: Prentice Hall).
Lowery, D. and Gray, V. (1995) ‘The population ecology of Gucci Gulch, or the natural regulation of
interest group numbers in the American states’, American Journal of Political Science, 39:1, 1–29.
Lyons, M. (2001) Third Sector: The Contribution of Nonprofit and Co-operative Enterprise in Australia (St
Leonards: Allen & Unwin).
Mitchell, N. (1990) ‘The decentralisation of business in Britain’, Journal of Politics, 52:2, 623–637.
Olson, M. (1965) The Logic of Collective Action (Harvard: Harvard University Press).
Oppenheim, F. (1975) ‘The language of political inquiry’, in F. I. Greenstein and N. W. Polsby (eds),
Political Science: Scope and Theory, Vol 1 (Reading, MA: Addison-Wesley).
Richardson, J. (1999) ‘Pressure groups and parties: a “haze of common knowledge” or empirical advance
of the discipline?’, in J. Hayward, B. Barry and A. Brown (eds), The British Study of Politics in the
Twentieth Century (Oxford: Oxford University Press).
Rose, R. (1985) Politics in England (4th edn) (London: Faber).
Salisbury, R. H. (1975) ‘Interest groups’, in F. I. Greenstein and N. W. Polsby, Nongovernmental Politics
Handbook of Political Science Vol 4 (Reading, MA: Addison-Wesley).
Salisbury, R. H. (1984) ‘Interest representation: the dominance of interest groups’, American Political
Science Review, 78:1, 64–78.
Salisbury, R. H. (1992) Interests and Institutions (Pittsburg: University of Pittsburg).
Sartori, G. (1970) ‘Concept misinformation in comparative politics’, American Political Science Review, 64:4,
1033–1053.
Sartori, G. (ed) (1984) ‘Guidelines for concept analysis’, in Social Science Concepts (Beverly Hills: Sage).
Schlozman, K. (1984) ‘What accent the heavenly chorus? Political equality and the American pressure
system’, Journal of Politics, 46: 1006–1032.
Schlozman, K. (1992) ‘Review of Walker, J., Mobilizing Interest Groups in America’, American Political Science
Review, 86.
Thomas, C. and Hrebenar, R. (1990) ‘Interest groups in the states’, in V. Gray, H. Jacob and R. Albritton
(eds), Politics in the American States (Scott, KS: Foresman/Little Brown).
Truman, D. B. (1951) The Governmental Process (New York: Alfred A. Knopf).
Walker, J. (1991) Mobilizing Interest Groups in America (Ann Arbor: University of Michigan Press).
Whorf, B. (1956) Language, Thought and Reality (Cambridge: MIT Press).
Wilson, D. (1983) The Lead Scandal (London: Heinemann).
Wilson, G. K. (1990) Interest Groups (Oxford: Basil Blackwell).
