0111-1760 University of Otago Economics Discussion Papers No. 0203
Policy Leadership Styles and the Process of Paradigmatic Policy Change: Three Propositions
Joe Wallis* and Brian Dollery** Abstract:
This paper formulates a theory of policy leadership based on propositions that relate to
the conditions under which rival leadership coalitions engage in a contest for authority over the
system-wide direction of the policy process and differentiate themselves according to distinctive
styles in respect of which the demand shifts due to the endogenous accumulation of
disappointment over distinct phases of a process of paradigmatic policy change. It both draws
from concepts familiar to policy theorists and the work of economic revisionists who have sought
to make the expressive dimension of phenomena such as leadership more amenable to deductive
: leadership styles. Policy paradigms, hope, disappointment.
Department of Economics
Department of Economics
School of Business
University of New England
Tel: (64) 3 479 8650
Tel: ++61 2 67 732500
Fax: (64) 3 479 8174
Fax: ++61 2 67 733280
2 Policy Leadership Styles and the Process of Paradigmatic Policy Change: Three Propositions Introduction
The view that government has as much to do with the problem of steering as it has to do with the
problem of power (Rose, 1987) has been taken up with a vengeance by contemporary writers on
the subject of "governance". They have attempted to move beyond the traditional focus on
decentralized/market-oriented and centralized/statist mechanisms of governance to highlight a
"third dimension" that explicitly blurs the boundaries between the public and private sectors and
involves state actors playing a catalytic role in engaging societal actors in network relationships
through which they strive to steer the policy process toward the realization of shared goals
(Rhodes, 1997; Stoker, 1998; Jessop, 1995). Within the fields of public administration and policy
studies this research represents an attempt to move away from the normative, formal,
constitutional understanding of the government as a "unitary state directed and legitimated by the
doctrine of ministerial responsibility" toward an attempt to understand the complex reality of
governing in practice where it is often the case that "there are many centers and diverse links
between many agencies of government at local, regional, national and supranational levels"
(Stoker, 1998, p.19). It thus has considerable affinities with the burgeoning literature on "policy
networks" that has significantly influenced implementation research in the field of policy studies
(Marsh and Rhodes, 1992).
In the view of Stoker (1998, p.18) the value of the emerging "governance paradigm" lies
"not at level of causal analysis" but rather rests in its capacity to provide an "organizing
framework", "a language and frame of reference" that leads theorists "to ask questions that might
not otherwise occur" regarding changing processes of governing. This paper advances the view
that the explanatory power of the governance paradigm could be enhanced if it is augmented with
a theory of policy leadership. This is based on the recognition that there is a striking affinity
between the way "governance" and "leadership" are conceived in the theoretical streams that
study these phenomena. This is reflected in the increasing tendency for governance theorists to
see "government as able to use new tools and techniques to steer and guide" (Stoker, p.24) while,
in modern leadership theory, the wide range of definitions of this phenomenon seem to be
converging toward the concept that "leadership is a social influence process through which the
members of a group are steered toward a goal" (Bryman, 1986, p.8). Moreover, within the long-
standing tradition of inquiry into leadership there appears to be a recurrence of a number of
concerns that are also strikingly relevant to governance theory. A comprehensive survey of this
literature by Bass (1990) suggests that these can be related to (a) the traits exemplified by leaders
(as identified in personal or "great man" theories of leadership) or shared in common by the
members of networks that collectively supply leadership in a particular context (Bryson and
Crosby, 1996; Wallis, 1999); (b) the conditions under which opportunities emerge for leadership
to play a historically significant role (as studied in "situational theories of leadership"); and (c) the
appropriateness of distinct leadership styles to different historical contexts or situations (Little,
This paper will attempt to formulate a theory of policy leadership within an institutional
context that addresses these concerns in a way that is directly relevant to contemporary
governance theory. This theory will be based on the following three propositions:
1. Policy leadership will be collectively supplied by that leadership coalition that is able to
prevail in a contest with its rivals for authority over the system- wide direction of the policy
2. The stability of the institutional context for policy making will determine whether policy
actors are concerned with the distinctive style as well as goals of the ruling leadership
3. The accumulation of disappointment with the style of leadership exercised by the ruling
leadership coalition will cause the demand for a new style of policy leadership to shift in a
predictable way during the distinct phases of a process of paradigmatic policy change.
In elaborating on the first two propositions, the paper will apply concepts derived from studies of
"policy networks", "advocacy coalitions" (Sabatier, 1991) and "policy paradigms" (Hall, 1993)
that have become familiar in the policy literature. The fact that many of the contributors to this
literature are economists and political scientists who are strongly influenced by the "rational
choice approach" does mean that they tend to describe rather than explain the expressive
dimension of leadership behavior. However, it is this dimension that would appear to underlie the
distinctiveness of leadership as a social influence process. Some pioneering work by "rational
choice revisionists" such as Elster (1998), Collins (1993) and Hirschman (1982, 1985) would
appear to make the expressive dimension of human behavior, in general, and leadership, in
particular, more amenable to deductive analysis. The paper will thus also draw on their work to
relate the internal cohesiveness of leadership coalitions and shifts in the emotional climate that
underlies the policy process to the cycles of hope and disappointment that are likely to occur
during the process of replacing one policy paradigm with another.
41. The Contest for Authority Between Rival Policy Leadership Coalitions
At any time the supply of policy leadership can be conceived as emerging from "a contest for
authority" between rival policy leadership coalitions (PLCs). The PLC that prevails in this contest
will be able to locate its members in key positions so that they can steer the policy process in the
direction of the bounded set of goals they share in common. It is proposed that these PLCs will
share a number of characteristics.
In the first place, the tasks of policy leadership undertaken by their members are typically
collectively supplied. A PLC can thus be conceived as a network of “policy entrepreneurs”
(Kingdon, 1984) who strive to advance one another into positions of leverage over the agenda-
setting, formulation, decision-making, implementation and evaluation stages of the policy
process. This concept of leadership ties in with much of the modern writing on the subject that
tends to emphasize the collective dimension of this phenomenon (Bryson and Crosby , 1992,
Secondly, its members include both state and societal actors drawn from across institutional
and partisan boundaries into horizontal leader-leader relationships similar to those observed in the
policy network literature. Their mutual concern is not, however, with the sectoral issues that
typically engage the participants in "policy subsystems" but with the system-wide direction of the
Thirdly, the contest for authority over the system-wide direction of governance may take a
form similar to that which Jenkins-Smith and Sabatier (1993) observe in subsystems that have
experienced the emergence of a stable line-up of opposing “advocacy coalitions”. These writers
propose that for this to occur the following conditions should be met: (i) the participants in a
policy subsystem should come to have a hierarchy of beliefs reflected in their unwillingness to
revise "policy core" as distinct from "secondary" beliefs in response to new information; (ii)
advocacy coalitions should come to be identifiable by the "policy core beliefs" which their
members share in common; (iii) the main controversies in a policy subsystem should involve
disputes about the core beliefs of opposing coalitions; and (iv) these disputes should typically not
be capable of uncontestable resolution through scientific methods or according to the standards of
independent professional forums but should tend to be perpetuated as each side buttresses its
position by using substantive policy information in an advocacy fashion.
The view that policy actors are identifiable in terms of their core beliefs suggests that
there is an expressive as well as an instrumental dimension to their behavior. As Charles Taylor
(1989) has pointed out, the commitments of scarce resources of time, wealth and attention that
people make to collective activities are not just based on an instrumental calculation of their
impact on the probability of the group realizing its goals (Olson, 1965) but on their motivation to
define their identity in a "public space" of questions about where they stand and who they
identify with by expressing through observable commitment how much these goals mean to them.
While policy theorists such as Sabatier have highlighted the significance of this type of
behavior, they have not made much progress in analyzing and explaining it, possibly because they
appreciate that this would involve them moving outside the boundaries of both the institutionalist
and rational choice traditions that shape their colleague’s understandings of political behavior.
There have, however, been some indications that these traditions are becoming more open to
explanations that take into account this expressive dimension.
In particular, a number of writers have sought to modify the rational choice perspective to
explain the effect the emotions might have on behavior. In his article "Emotions and Economic
Theory" (1998), Elster proposes that emotions are triggered and sustained by beliefs that are
expressed with an observable level of emotional energy so as to produce "action tendencies" or,
as Frijda (1986, p.70) put it, “states of readiness to execute a given type of action”. Elster rejects a
cost-benefit model of the emotions that treats them "as psychic costs and benefits that enter into
the utility function on a par with satisfactions derived from material rewards" (1998, p.64) in
favor of an approach that views them both as sources of "cognitive dissonance" (Festinger, 1957)
and as mechanisms of dissonance reduction. In his view individuals do not choose emotions
since their occurrence is “basically unbidden”. A dissonance model of the emotions could,
however, explain why they choose to avoid or seek out situations that are likely to trigger certain
Randall Collins (1993) has followed a strikingly similar line of argument to explain how
interactions of a sufficient "density" between the members of a group that hold in common beliefs
with an “emotional energy” that is "empirically visible, both in behavior (especially nonverbal
expressions and postures) and in physiology" (p.211) can cause the participating group's focus of
attention and common emotional mood to go through a short term cycle of increase and mutual
stimulation until a point of emotional satiation is reached. According to this writer, these
interactions will leave each participant with an "energetic afterglow" that "gradually decreases
over time" so that individuals have an incentive to reinvest their emotional energy in subsequent
interactions. It may therefore accumulate across interactions so that "an individual may build up
a long-term fund of confidence and enthusiasm" (p.212).
These lines of thought can be applied to explain the formation and internal cohesion of
PLCs if these networks are seen as providing the context within which the shared hopes
members can be strengthened. Snyder's (1994) definition of hope as "the sum of the willpower
and waypower that you have for your goals" (p.5) suggests that it can be associated with an action
tendency to keep striving, in the face of repeated disappointments, to advance particular goals.
This source of motivation would seem to be strikingly relevant to the behavior of the members of
PLCs. As they strive to advance their goals they would typically encounter resistance from other
groups and from the institutional and environmental factors that constrain their capacity to "get
their way" and generate the stream of disappointments that can accumulate in a way that weakens
their political resolve. The corrosive effect of disappointment on this action tendency may,
however, not be immediately apparent since the members of PLCs may make an allowance for
disappointment up to a threshold determined by the strength of the beliefs they hold about its
Two core beliefs, in particular, would seem to underlie their action tendency to keep
striving to advance their PLC's goals. The first is the belief that the advancement of these goals is
"neither impossible nor inevitable" (Sutherland, 1989, p.195). This belief does not have to be
based on probabilistic calculation - it may only be derived from an "imagined skein of
possibilities" (Shackle, 1973, p.62). For the members of a PLC it may be sufficient that they
believe that they have the "waypower" (Snyder, 1994) to effectively react to obstacles and
resistance by devising and pursuing alternative ways to advance their collective goals.
The second belief is that the collective advancement of these goals is "worthwhile" or
"important" in the sense that it is "worthy of pursuit in a special way incommensurable with other
goals we might have" (Taylor, 1985, p.135). The process of placing hope in certain goals seems
to involve an investment or commitment of self to the realization of these goals. Or, to use
Hirschman's (1982) terminology, it requires actors to form a "second order metapreference"
regarding the "kind of life they want to live" or the "kind of person they want to become".
Elster's proposed "dissonance theory of the emotions" (1998) suggests that this type of
actor will seek out situations in which the dissonant effect of cumulative disappointment can be
countered and the beliefs underlying hope can be strengthened. In particular, their quest for two
types of cognition will draw them to interact within PLCs that are bounded by shared "core"
beliefs. Firstly, the reasons individual members have for striving to advance the goals of the PLC
will always, to a degree, be implicit, inchoate and partly articulated. They will therefore look to
one another to provide a clearer, more explicit articulation and to buttress their beliefs in the
worth and possibility of their collective leadership. Secondly, the emotional energy that is
produced and reproduced in these interactions in the manner described by Collins (1993) will
augment the fund or reserve of "willpower and waypower" (Snyder, 1994) that they need to draw
on if they are to keep striving to overcome and circumvent the institutional and political obstacles
to the advancement of their goals.
It should be emphasized that these benefits do not accrue to "free-riders" or "preference
falsifiers" (Kuran, 1990). A person who does not genuinely hold the core beliefs of a PLC will
find it difficult to "keep up an act", continuously "fooling" other members about their lack of
passionate intensity and, even if they succeed in this falsifying strategy, they will derive no
satisfaction from a sense of belonging to this group. The internal cohesion of PLCs may
therefore increase over time as they screen out those participants who cannot derive "solidary in-
process benefits" (Buchanan, 1979) from interacting with other members who share their beliefs
and invest an observable level of emotional energy in these interactions.
It would therefore seem plausible that through these mechanisms a number of internally
cohesive PLCs would seek to engage, at any time, in the contest for authority over the direction of
policy development. The question of how these PLCs differentiate themselves from one another
must now be explored in more detail.
2. Policy Leadership Coalitions in Punctuated Equilibrium Models of Policy Change
Perhaps the most distinctive aspect of the "new institutionalist" approach to policy studies is its
attribution of "historical inefficiency" to the failure of the institutions of policy making fail to
adapt rapidly to changes in the policy environment so that institutional change tends to follow a
"punctuated equilibrium" pattern. PLCs can be distinguished according to their tolerance for the
inefficiency of history. Punctuated equilibrium models suggest this tolerance will be pervasive
during the comparatively long periods when PLCs operate within the boundaries of what Hall
(1993) terms a ruling "policy paradigm".
During these periods most PLCs will exercise a "political" style of leadership that will
focus on realizing particular goals without seeking to change the institutional context. According
to Hall (1993) the policy changes they initiate will tend to be either "second order" changes in
policy instruments and "first order" adjustments in the settings of these instruments. They will
typically avoid advancing the process of replacing one paradigm with another, a process Hall
characterizes as involving a "third order" change in the hierarchy of policy goals and the
overarching terms of policy discourse.
Hall cites the shift from Keynesian to monetarist macroeconomic policies under the
Thatcher government in the UK as an example of third order change. However, this order of
change is also apparent in those developing and transitional countries that have, over the last two
decades, implemented comprehensive reform programs (CRPs) based on the “Washington
consensus” that recommends the abandonment of Keynesian demand management and import-
substituting industrialization policies in favor of a strategy that focuses on "macroeconomic
stabilization" (of debt and inflation) and "structural adjustment" through market-oriented reforms
" (Williamson, 1994; Rodrik, 1996).
In a survey of the political economy of policy reform, Rodrik (1996, p.10) has suggested
that the main issue confronting the contributors to this literature relates to the question: "Why are
so many governments reforming now, after decades of adherence to policies of the opposite
kind?" Generally, it would seem that the delayed implementation of CRPs can be related to the
risks they pose for governing coalitions. During periods of "normal politics" (Balcerowicz, 1994)
they would seem to have propensity to avoid the political risks of radical shifts that are
surrounded by ex ante
uncertainty about their distributional consequences (Rodrik, 1996), that
depart from the centrist position associated with the Downsian consensus and that may create an
opportunity for a coalition of minorities opposed to comprehensive reform to win the next
election (Wallis and Dollery, 1999, p.184). The leadership style exhibited by most PLCs during
periods of paradigm stability will thus be pragmatic and incrementalist as they engage in political
processes of bargaining and deliberation to advance their goals within a largely given institutional
Hall (1993) argues that the authority of the reigning policy paradigm will gradually eroded
by the accumulation of "anomalies" and the resort by policymakers to "ad hoc experimentation"
that stretches its coherence. I would argue that this process will be accompanied by an
accumulation of disappointment with the prevailing political leadership style. In many ways this
is analogous to the accumulation of specific disappointments with particular lifestyles that,
according to Hirschman (1982), was a significant endogenous factor precipitating shifts in the
percentage of a population that exclusively engaged in private pursuits, on the one hand, and
committed themselves to activist public causes, on the other.
The accumulation of disappointment with the failure of PLCs to offer alternatives to a
political style of policy leadership will eventually produce a climate of frustration with a policy
community. A growing number of policy actors will experience frustration with the
unwillingness of PLCs to question, challenge or consider changes outside the boundaries of the
ruling paradigm and associated institutional framework. This frustration may be countered by the
emergence of PLCs that differentiate themselves not just in terms of their goals but also in terms
of their capacity to exercise a style of leadership that makes an effective contribution to the
advancement of a paradigmatic "third order" process of policy change that restores hope by
overcoming the institutional obstacles to the realization of these goals.
But what are the alternatives to the "political" style of policy leadership? Graham Little
(1988) suggests three. In the first place, there is the “inspirational” style exhibited by leaders who
tend to be "political Pandoras, liberating hopes . .unrealistic, inventive imprudent, careless,
enraptured with change and the future" (1988, p.5). Little suggests that at certain stages of their
political careers John F. Kennedy in the United States, Harold Wilson in the United Kingdom,
Pierre Trudeau in Canada, Gough Whitlam in Australia and David Lange in New Zealand may
have exhibited some of these traits. Vaclav Havel in the Czech Republic and Nelson Mandela in
South Africa would seem to be more recent exemplars of this style of leadership.
Secondly, there is the “strong” style exhibited by leaders who "prefer to implement ideas
rather than to debate them" (Little 1988, p.45). Although they are “deliberately unvisionary and
unexciting” (p.5) they have a reputation for decisive action based on "simple, tangible goals,
minimal entanglements and reluctance to compromise" (p.15). Little devotes much of his book to
examining the degree to which Margaret Thatcher in the United Kingdom, Ronald Reagan in the
United States and Malcolm Fraser in Australia conformed to this type.
Thirdly, there are leaders who exercise what Little calls a “group” style of leadership.
They “are reluctantly aggressive and tend to idealize solidarity, equality and consultative
processes” (p.6). Little tends to see them as more appealing but less effective than strong leaders.
In this regard he argues:
"President Carter is an outstanding example from this period as is Michael Foot, the former British
Labour Leader. Reagan beat one, Thatcher the other, as they did their successors. Mondale was a
classic Group Leader in his attachment to the solidarities of working men and women and his
preachments on compassion. ‘Sunny’ Jim Callaghan, fruitlessly searching for peace in industrial
relations, went under the firmer bite of Mrs Thatcher. In Australia, an exemplary action by Bill Hayden
brought Labor to power in 1983. Hayden, a Group Leader, resigned to make way for Hawke ‘for the
good of the Party’, and Hawke went on to beat Fraser" (p.6).
I would suggest that this conclusion is dependent on the context of 1980s politics that Little is
studying and fails to appreciate the effectiveness of the style of leadership offered in the 1990s by
leaders such as Bill Clinton in the US and Tony Blair in the UK. This style can be more helpfully
characterized as “empathetic” since while it encompasses the more inclusive style that Little
associates with Group Leaders, it dispenses with the nostalgia that can make this style of
leadership irrelevant in the aftermath of radical change and emphasizes the necessity of “adjusting
to the new realities”.
Little tends to view these styles of leadership as being exercised by particular individuals.
To relate his typology to the concept of policy leadership being advanced in this paper, it will be
necessary to delineate the type of network through which each style will be collectively supplied
and the distinctive "expression games" played by the members of these networks to differentiate
their style from that of their rivals. This will lay the basis for elaborating the third proposition
that the demand for leadership styles will shift in a predictable way during the distinct phases of a
process of paradigmatic policy change. 3. Expression Games and the Path of Leadership Succession
During those periods of institutional instability when PLCs seek to differentiate themselves in
terms of alternative leadership styles, they will engage in what Goffman (1959) described as
"expression games". These are typically a form of social interaction that involves "senders" who
express themselves in particular ways, and "receivers" who take in and react to such
expressions, forming an impression of the "senders". This concept is particularly pertinent to
policy studies since the interpretation of political expression generally involves "making
inferences from the expressive act about the sender's motives, values and commitments" (Loury,
1994: 432-3). Through repeated plays of an expression game a PLC can construct a stable
impression of the style of leadership its members are striving to supply.
As Table 1 indicates the boundaries of the policy quest the suppliers of a particular
leadership style are striving to advance may be defined by differentiating their leadership style
from that of its two most relevant alternatives. Table 1: The Expression Games Associated With Different Styles of Policy Leadership Style Policy Network Different- To create Irrelevant quest Focus iated from impression of alternative . . . . . .
• Autonomy from
old rules, roles
Inspirational Innovation Formul-
• Open-ness to
• Commitment to
coherent set of
• Resistance to
• Concern with
losers to adjust
to new realities
• Discontent with
It is proposed that inspirational leadership will tend to be supplied by PLCs that are engaged in a
quest for policy innovation. One person, the "inspirational leader" is likely to play an even more
focal role within these networks than in the case of those associated with other styles of policy
leadership. Inspirational leaders tend to present themselves as alternatives to leaders with a
- Policy Leadership Styles and the Process
- of Paradigmatic Policy Change: Three Propositions
- Joe Wallis* and Brian Dollery**
- Blair, T. (1998). Leading the Way: A New Vision for Local Government. London: IPPR.
- Eichbaum, C., (1999) The Politics and Economics of the Third Way. In S. Chatterjee (ed.) The New Politics: A Third Way for New Zealand. Palmerston North: Dunmore Press.
- Greenwald, B. and Stiglitz, J. (1993). New and Old Keynesians. Journal of Economic Perspectives, 23-44.
- Kydland, F. and Prescott, E. (1977) Rules Rather Than Discretion: The Inconsistency of Optimal Plans. Journal of Political Economy 85(3): 473-491.
- Wilson, J. (1989). Bureaucracy: What Government Agencies Do and Why They Do It. New York: Basic Books.