The Conduct of Policy Analysis: Philosophical Points of Reference

21
559 The Conduct of Policy Analysis: Philosophical Points of Reference John Dixon and Rhys Dogan University of Plymouth Abstract This paper explicates and reviews the competing approaches to policy analysis. It does so by construct- ing a methodological taxonomy that enables the identification of competing philosophical methodologies that underpin contending approaches to policy analysis. This is done by reference to contesting under- standings of what constitutes knowledge (an epistemological issue) and what exists that is capable of giving rise to consequences (an ontological issue). This philosophical framework is then, illustratively, applied to an examination of the concept of the underclass. Then the epistemological and ontological challenges facing policy analysts seeking to analyze and address policy problems and issues are identi- fied. Meeting these challenges requires the adoption of a methodology that draws insights from the epis- temological and ontological syntheses that have emerged within contemporary social theory. Introduction This paper explicates and critically reviews, within a philosophical framework, the competing approaches to policy analysis. It does so by drawing upon the philoso- phy of the social sciences to construct a taxonomy of contesting epistemological and ontological understandings. With the resultant set of epistemological and ontolog- ical configurations (or methodological families) come a set of attendant axiomatic propositions with respect to ethics and values, human nature and motivation, the meaningful interpretation of valid information, decision analysis, power and com- pliance, and salient analytical risks. This enables the articulation of contending approaches to how the causes and consequences of policy problems or issues should be described and analyzed; and how policy solutions should be conceptualized, assessed, and judged, so enabling the enunciation of the salient analytical risks asso- ciated with policy analysis that only draws upon one of these contending method- ological families. This philosophical framework is then, illustratively, applied to an examination of the concept of the underclass. Then, the epistemological and onto- logical challenges facing such policy analysts are identified. Meeting these chal- lenges requires the adoption of a methodological approach that draws insights from the epistemological and ontological syntheses that have emerged within contem- porary social theory. A Philosophical Framework for the Analysis of Policy Any analysis of policy is embedded in particular contending and incompatible judg- ments about the ultimate constituents of social reality and how they can be known and understood. This leads to incomplete, not to say incompatible, analytical out- comes, contending perspectives on problem causation and problem solution. Thus, how policy analysts choose to interrogate the social word, so as to describe, explain, understand, judge and address policy problems and issues depends on their philo- Review of Policy Research, Volume 21, Number 4 (2004) © 2004 by The Policy Studies Association. All rights reserved.

Transcript of The Conduct of Policy Analysis: Philosophical Points of Reference

Page 1: The Conduct of Policy Analysis: Philosophical Points of Reference

559

The Conduct of Policy Analysis: Philosophical Points of Reference

John Dixon and Rhys DoganUniversity of Plymouth

Abstract

This paper explicates and reviews the competing approaches to policy analysis. It does so by construct-ing a methodological taxonomy that enables the identification of competing philosophical methodologiesthat underpin contending approaches to policy analysis. This is done by reference to contesting under-standings of what constitutes knowledge (an epistemological issue) and what exists that is capable ofgiving rise to consequences (an ontological issue). This philosophical framework is then, illustratively,applied to an examination of the concept of the underclass. Then the epistemological and ontologicalchallenges facing policy analysts seeking to analyze and address policy problems and issues are identi-fied. Meeting these challenges requires the adoption of a methodology that draws insights from the epis-temological and ontological syntheses that have emerged within contemporary social theory.

Introduction

This paper explicates and critically reviews, within a philosophical framework, thecompeting approaches to policy analysis. It does so by drawing upon the philoso-phy of the social sciences to construct a taxonomy of contesting epistemological andontological understandings. With the resultant set of epistemological and ontolog-ical configurations (or methodological families) come a set of attendant axiomaticpropositions with respect to ethics and values, human nature and motivation, themeaningful interpretation of valid information, decision analysis, power and com-pliance, and salient analytical risks. This enables the articulation of contendingapproaches to how the causes and consequences of policy problems or issues shouldbe described and analyzed; and how policy solutions should be conceptualized,assessed, and judged, so enabling the enunciation of the salient analytical risks asso-ciated with policy analysis that only draws upon one of these contending method-ological families. This philosophical framework is then, illustratively, applied to anexamination of the concept of the underclass. Then, the epistemological and onto-logical challenges facing such policy analysts are identified. Meeting these chal-lenges requires the adoption of a methodological approach that draws insights fromthe epistemological and ontological syntheses that have emerged within contem-porary social theory.

A Philosophical Framework for the Analysis of Policy

Any analysis of policy is embedded in particular contending and incompatible judg-ments about the ultimate constituents of social reality and how they can be knownand understood. This leads to incomplete, not to say incompatible, analytical out-comes, contending perspectives on problem causation and problem solution. Thus,how policy analysts choose to interrogate the social word, so as to describe, explain,understand, judge and address policy problems and issues depends on their philo-

Review of Policy Research, Volume 21, Number 4 (2004)© 2004 by The Policy Studies Association. All rights reserved.

Page 2: The Conduct of Policy Analysis: Philosophical Points of Reference

sophical disposition, the constituents of which are their epistemological pre-disposition (contentions held about what is knowable, how it can be known, andthe standard by which truth can be judged) and their ontological predisposition(contentions held about the nature of being, what can and does exist, what the con-ditions of existence might be, and to what phenomena might causal capacity beascribed).

The Contending Epistemological Dispositions

Epistemology investigates the relative veracity of knowledge claims. It is concernedwith evaluating the relative reliability of explanations based on different methodsof knowledge acquisition, thereby critiquing the knowledge produced by diverseknowledge systems (Pitt, 1989). The conversion of a knowledge claim into genuineknowledge requires a criterion or standard by which judgments can be made aboutwhat constitutes sufficient justification to warrant a knowledge claim being treatedas a justified knowledge claim, or the truth (Moser, 1989).

There are two broad approaches to epistemology offered by the philosophy of the social sciences. The first is naturalism, which uses inductive and deductive rea-soning to explain social phenomena. Knowledge claims, therefore, can or shouldonly be justified by reference to individual experience or reason. The second ishermeneutics, which proposes that human beliefs are the final arbiter of social reality.Knowledge claims, therefore, can or should only be justified by reference to testimony, expert authority, consensus, common sense, or received wisdom (Hollis,1994). These distinctive epistemological approaches are reflected in a variety ofdichotomous terminology: explanation versus understanding, objectivism versussubjectivism, materialism versus idealism, or empiricism versus constructivism.

Naturalism

Naturalism assumes that the social world is objective and material. The truth aboutit should thus be grounded in the observation and analysis of material phenom-ena. Naturalist epistemology proposes two types of knowledge. Synthetic statements,which contain a posteriori (empirical or sensory experience-based) knowledgederived through inductive logic (Hempel, 1966), describe relationships in theknowable social world identified through observation. The veracity of these knowl-edge claims rest on the probability of previous patterns of observation beingrepeated in the future. Analytic statements, which contain a priori (nonempirical)knowledge derived from deductive logic, present syllogistic reasoning or self-evident tautologies, which means that they can be logically true but empiricallyuntrue (Williams & May, 1996, p. 25). While both the inductive and deductiveforms of naturalism address the problem of causality in different ways (Popper,1935/1959/2000), they offer only reasonably reliable predictions, for they cannotidentify unambiguous causal relationships (Williams & May, 1996).

There are many epistemological positions to be found under the rubric of naturalism, including, inter alia, empiricism (stressing the fundamental role ofexperience as the foundation of all knowledge) (Hume); verificationism (verifyingknowledge claims involves empirical or observational evidence for believing thatthey are true) (Hempel); logical positivism (arguing that deductive logic and induc-

560 John Dixon and Rhys Dogan

Page 3: The Conduct of Policy Analysis: Philosophical Points of Reference

tive inference is the foundation of knowledge) (Schlick); and falsificationism (verify-ing knowledge claims involves systematically attempting to refute or falsify themby use of empirical or observational evidence) (Popper).

Hermeneutics

Hermeneutics assumes that the social world is subjective and ideational (Baert, 1998;Hollis, 1994). The truth about it should, thus, rest on interpretations embeddedin day-to-day expressions derived from practice, discourse, and language (Winch,1990). Human knowledge is, therefore, generated by acts of ideation (personalreflections and ruminations) that rest on intersubjectively shared symbols, or typifications, that allow for the sharing (reciprocity) of perspectives (Schutz,1932/1967). However, this pre-acquaintance of the individual with the social is bothactive and indexical (contextually variable, depending on the relationship betweenperception, memory, and communications that connects thought to object) andrequires acts of reflexive interpretation to ensure the appropriate meaning con-textualization (Blumer, 1969; Garfinkel, 1967). Thus, reasoning and morality arerelative across time, societies, and individuals (epistemic and moral relativism, respec-tively), thus, people in different societies have very different beliefs and beliefsystems (social relativism). Hermeneutic knowledge is, thus, not only culturally spe-cific and subject to severe relativism, which makes explanation contingent, but alsodynamic and open to constant revision, which makes prediction problematic (Hollis& Lukes, 1982).

There are many epistemological positions to be found within hermeneutics, includ-ing, inter alia, epistemological hermeneutics (discovering meaning that is the con-struction of individual subjectivity through comparison and contrast with meaningconstructed by others) (Schleiermacher and Dilthey); existentialism (investigating thenature of human existence by focusing on individuals in their search for their senseof who they are and for an authentic way of life or self-fulfillment) (Kierkegaardand Sartre); phenomenology (investigating the relationship between the objective andsubjective on the basis of the processes by which people identify distinct and mean-ingful experiences in their world, as constituted in and by their consciousness)(Husserl and Heidegger); and linguistic epistemology (learning the rules of languagethat shape knowledge acquisition because they are intersubjective shared androoted in a taken-for-granted stock of knowledge) (Wittgenstein and Winch).

The Contending Ontological Dispositions

Ontology seeks to provide a coherent account of the nature and status of the cat-egories of phenomena that possess causal efficacy. It considers the veracity and reli-ability of explanations based on different notions of causality. In the social sciences,the ontological debate is on the relationship between the external (structuralism) andinternal (agency) dimensions of human behavior (Wendt, 1999).

Structuralism

Structuralism contends, in the tradition of Comte, that social action is caused bysocial structures—“the ordered social interrelationships, or the recurring patterns

The Conduct of Policy Analysis 561

Page 4: The Conduct of Policy Analysis: Philosophical Points of Reference

of social behavior that determine the nature of human action” (Parker, 2000, p. 125)—which means that individuals are merely indeterminate matter. Thesesocial structures, which are external to individuals, “mould [constrain] people’sactions and thoughts” (Baert, 1998, p. 11). It is axiomatically taken to be difficult,if not impossible, for one person to transform them (Baert, 1998, p. 11).

There are many ontological positions to be found within structuralism, including,inter alia, anthropological structuralism (explaining how members of a society relateto social organizations and societal structures) (Levi-Strauss); functional-structuralism(explaining a social action or the existence of a social phenomenon in terms of the contributions made to the continuity of a society) (Durkheim, Merton, andParsons); historical materialism (explaining how particular forms of society responsi-ble for observed social phenomena come into existence by reference to historicalsocioeconomic processes and relations) (Marx); linguistic structuralism (meaningabout the social world and making sense of what others say is produced by language, as a set of rules that govern the combination of sounds, thus meaning[content-elements] cannot be identified independently of sounds and words[expression-elements]) (Saussaure); symbolic interactionalism (meaning about thesocial world [everyday life] and making sense of what others say emerges throughthe interaction of language, identity, and society, the product of the interactiveprocesses that underpin the construction, negotiation, presentation, and affirma-tion of self (symbolic communications) (Blumer); language games (meaning aboutthe social world and making sense of what others say is the product of the complexof interwoven speech and action that are ordered in praxis in a way that makessense to the participants) (Wittgenstein); poststructuralism (meaning about the socialworld is derived from self-reflexive discourses that acknowledge the inherentlyfragmented, diverse, tenuous, ambiguous, and culture-specific nature of knowl-edge, which is always changing and contestable, and never final and complete)(Derrida and Foucault); symbolic interactionalism (meaning about the social world[everyday life] and making sense of what others say emerges through the interac-tion of language, identity, and society, the product of the symbolic communicationsamong interacting social actors, which emphasize the interactive processes thatunderpin the construction, negotiation, presentation, and affirmation of self)(Mead); and postmodernism (meaning about the social world can only be situational,provisional, contingent, and temporary, because “notions of truth, objectivity,reason and other grand narratives have been superseded by notions of diversity,instability, fragmentation and indeterminacy” (Nightingale & Cromby, 1999, p. 228) (Baudrillard and Blanchot). Structuralism’s dilemma is that it might be ableto explain the empirically strong correlation between individual behavior and socialcohort, but it cannot unambiguously explain outliers derived from acts of choiceby free individuals (Williams & May, 1996).

Agency

Agency contends that social action can only be caused by individual intention. Inthe tradition of Hegel, Hursserl, and Schutz, it refers to explanations of humanbehavior that focus on individuals as voluntaristic human agents who ascribe sub-jective meaning to their actions. Its central proposition is that individuals, enabled

562 John Dixon and Rhys Dogan

Page 5: The Conduct of Policy Analysis: Philosophical Points of Reference

by their psychological and social psychological make-up (Parker, 2000), activelyinterpret their social reality, and, in the light of their mental states (say, hope, belief,and desire) they then focus their minds on objects, and act accordingly (Baert,1998). For people to be moved to action thus requires not only a set of beliefs aboutthe way things are, about the actions possible given the way things are, and aboutthe likely effects of those actions on the way things are, but also a desire to changethe way things are by a course of action in the hope that it brings about the intendedchange (Marks, 1986).

There are many ontological positions to be found within agency, including, interalia, rational choice theory (explaining the actions of individuals by reference to theirown individual self-interest motivations) (Hobbes, Machiavelli, Manderville, Smith,Arrow, and Downs) and game theory (making rational decisions under conditions ofuncertainty) (von Neumann and Morganstern); social phenomenology (meaning aboutthe social world and making sense of what others say is individually constructed byusing meaning contexts that reflect patterns of subjective meaning acquiredthrough intersubjective social interactions) (Schutz); dramaturgical analysis (explain-ing how, by reference to theatrical metaphors, people in their social interactionsstrive to present an image of themselves in particular ways) (Simmel and Goffman);and ethnomethodology (meaning about the social world and making sense of whatothers say is individually constructed by reference to actual, ordinary activities[practical actions, practical circumstances, commonsense knowledge of social struc-tures, and practical reasoning]) (Garfinkel). Agency’s dilemma is that it can appar-ently explain the empirically strong correlation between individual behavior andfree choice, but it cannot explain outliers that are the product of ideational con-siderations (Williams & May, 1996).

The Contending Methodological Families

These epistemological and ontological dichotomies give rise to a set of four configurations of epistemological and ontological assumptions—philosophical dispositions (methodological families). Each establishes distinctive, philosophicallycoherent methodologies that can be used to describe, explain, understand, judgeand change the social world, thereby providing a coherent set of inquiry agendas,theories, and methods of investigation (Dixon 2003; Dixon & Dogan, 2002, 2003a,2003b, 2003c, 2003d). These methodological families are captured in Figure 1.

A Philosophical Taxonomy of the Approaches to the Analysis of Policy

Each of the methodological families identified in Figure 1 supports a coherent setof approaches to the analysis of policy.

The Naturalist-Agency Perspective

Policy analysts predisposed to a naturalist-agency philosophical stance would con-sider the social world to be a knowable objective reality, which they would charac-terize as an aggregation of individuals, each of whom interact, exercise theirfreedom of choice, and establish contractual relationships. As they hold that indi-

The Conduct of Policy Analysis 563

Page 6: The Conduct of Policy Analysis: Philosophical Points of Reference

viduals are self-determining, people are presumed to conduct their affairs in asocial context where no one has a preordained place and their commitment is onlyto themselves: “I decide what I will think.”

Ethics and Morality—Naturalist-agents would generally accept the Kantian proposi-tion of the right to self-determination of rational agents. This means that rationalagents have the capacity to subject their desires and values to rational scrutiny asa precursor to exercising self-determination. Rational agents should, thus, bebound only by self-given moral principles that originate in the exercise of reason(Christman, 1989; Dworkin, 1988). They would accept the contractarian proposi-tion that moral principles can be construed in terms of rational bargaining amongself-interested individuals (Hampton, 1986), which can be deduced from the ratio-nal choices made, behind a veil of ignorance, by self-interested contractors deprivedof information about their talents and attributes (Rawls, 1971).

Axiologically, naturalist-agents are nondogmatic ethical realists seeking to respondto ethical problems by means of scientific enquiry. In essence, they would hold thatthe rightness of an action can be judged by reference either to moral facts ( justifiedmoral knowledge claims) grounded in some nonmoral position (such as individualpleasure or utility), or to rules for moral action, both of which determine what is rightaction (moral realism) (Brink, 1989). Their preferred solution to moral conflict, then,would involve assessing the net beneficial consequences of an action (Gouinlock,1972; Meyers, 1986). Thus, they would prefer to judge the rightness of people’s

564 John Dixon and Rhys Dogan

Epistemology Naturalism

Structuralism

Naturalist Structuralism: Presumes an objective social world, knowable by the application of scientific methods, in which social structures exercise power over agency, which makes human behavior predictable. Embraces, inter alia, anthropological structuralism, linguistic structuralism, functional structuralism, and historical materialism.

Hermeneutic Structuralism: Presumes a subjective social world, knowable only as it is socially constructed, with people’s action being determined, and made predictable, by their collective interpretation of this reality. Embracing, inter alia, symbolic interactionalism, language games, post-structuralism, and post-modernism.

Ontology

Agency

Naturalist Agency: Presumes an objective social world, knowable by the application of scientific methods, in which people are agents of their actions, with their behavior made predictable by their unconstrained self-interest. Embraces, inter alia, rational choice theory, and game theory.

Hermeneutic Agency: Denies the objectivity of social reality, which is only contestably knowable as what people believe it to be, with agency constrained by their subjective perceptions of social reality, which makes human behavior unpredictable. Embraces, inter alia, social phenomenology, dramaturgical analysis, and ethnomethodology.

© John Dixon and Rhys Dogan, 2004.

Hermeneutics

Figure 1. A Taxonomy of Philosophical Methodologies

Page 7: The Conduct of Policy Analysis: Philosophical Points of Reference

actions either by the value of their actual, or even intended, effects (ethical conse-quentialism) in terms of producing the most good (act-consequentialism, or, in utili-tarian terms, act-utilitarianism), even if this is not their intention; or by whether theyfollow a set of rules that are generally accepted to be the best way to promote themost good (rule-consequentialism or, in utilitarian terms, rule-utilitarianism) (Scarre,1996; Scheffler, 1988). Right action is, thus, grounded in some conception of finalpurpose or good—the principle of ultimate utility (teleological ethics [Muirhead,1932])—hence, the end justifies the means, which makes such an action intrinsi-cally good (Bentham, 1789/1970; Dewey, 1932/1985; Mill, 1863/1968). In coming tothis conclusion they could well place morally decisive importance in a particularset of circumstances (situational ethics) on the actual the consequences achieved(empirical situationism) or on improving matters in ways a given a situation, with itsdistinctive constraints, makes viable (mournful realism) (Fletcher, 1966).

Human Nature—The naturalist-agents’ metaphysical explanation of what it is to bea human being draw upon the traditions of liberalism (individuals are essentiallyfree beings) (Bentham, Locke, Mill, and Rousseau); libertarianism (individuals areseeking liberation from the interference of other people) (Hayek and Norzick); util-itarianism (individual conduct is right or wrong because of its tendency to producefavorable or unfavorable consequences for the people who are affected by it)(Bentham and Mill); and hedonism (individuals pursue their own pleasure at alltimes) (Bentham, Epicurus, Mill, and Sidgwick). To naturalist-agents, there is a uni-versal human nature that can be characterized as self-serving and selfish, whichmeans that people’s behavior can be manipulated by material incentives and dis-incentives. Who a person becomes, then, is contingent upon how that personexploits.

Justified Knowledge—To naturalist-agents any statement that cannot be verifiable, totheir satisfaction, by means of abstract thinking (deductive reasoning) or the sci-entific method (inductive reasoning) would be relegated the status of expressionsof irrational emotions, subjective values, or beliefs, and their validity would bedenied.

Rationality Mode—Naturalist-agents would prefer to interpret information by theapplication of a formal, synoptic mode of rationality that focuses on the achieve-ment of purposes, goals, and ends (intrinsic teleology) (Woodfield, 1976). Thisapproach to interpretation applies practical reasoning (O’Neill, 1989), which is con-cerned with applying deductive logic or inductive inference to identify ways ofdetermining not only which problems are solvable but also which goals are worthconsidering. Thus, they would subscribe to the belief that the ends should be unex-ceptionable (such as maximizing profit and utility) and so crucial that theyinevitably justify the means used to achieve them. This avoids the problem of resolv-ing contestable objectives, which would delay action by initiating pointless valuesdiscourses. They would prefer problems to be defined in a highly structured(reductionist) way, which would facilitate causal and evaluative information gathering.

The Conduct of Policy Analysis 565

Page 8: The Conduct of Policy Analysis: Philosophical Points of Reference

Decision Analysis—Decision analysis for naturalist-agents would involve selectingfrom among all alternative means (strategies) the one that creates the best oroptimal opportunity for improvement (perhaps conceived as the most cost-beneficial, cost-effective, or cost-efficient).

Power and Compliance—Naturalist-agents prefer to exercise Boulder’s (1990)exchange power, French and Raven’s (1959) resource or reward power, or Hales’s(2001) economic power. Thus they have a preference for rules specifying commit-ments, promises, and offerings that entail rights and duties (commitment rules [Onuf,1989]). Instrumental compliance would be expected, based on an economic calcu-lation of the compliance costs and benefits involved (Etzioni, 1961).

Salient Analytical Risks—Naturalist-agents face the salient analytical risks of beingunable to understand either the nature, consequences, and causation of policyproblems or issues because they cannot be explained only by the application of nat-uralist methods or because they cannot accommodate behavior that is induced byideational considerations (such as ideological or religious obligations), or why theirsolutions, which presume an agency ontology (involving individual incentives anddisincentives) is ineffective because of uncertainty, asymmetrical information,opportunism, and unenforceable contracts.

Naturalist-Structuralist Perspective

Policy analysts predisposed to a naturalist–structuralist philosophical stance wouldconsider the social world to be a knowable objective reality, which they would char-acterize as a (hierarchical) social order based on positional authority, expressedthrough orderly differentiation (Dumont, 1970). These social structures establishin those affiliated with them a sense of identity. They would thus presume thatpeople conduct their affairs in a social order in which everyone has, and knows,their preordained place, although that may well vary over time—they decide whatwe should think.

Ethics and Morality—Naturalist-structuralists would adhere to the Kantian deonto-logical ethical proposition that ethical principles are discoverable by reason. Thisgives rise to moral (categorical) imperatives (generalized moral judgments thatexclude self-interest and that a rational moral agent would follow), which makecertain kinds of obligations or duties intrinsically right (Kant, 1781/1998, 1785/2002,1788/1996). Ethical propositions—made de rigueur by trusted authority figures—that emphasize duty and obligation would, thus, be seen as justifiable reasons forsocial action.

Human Nature—The naturalist-structuralists’ metaphysical explanation of what it isto be a human being would draw upon the tenets of classical realism (individuals arerational beings seeking pleasure, happiness and goodness) (Aristotle, Socrates, andPlato) and idealism (individuals are essentially rational moral beings with the capac-ity to rise above their feelings and passions) (Kant). People may well be errant, butthey are redeemable by good institutions that demand conformity with the norms

566 John Dixon and Rhys Dogan

Page 9: The Conduct of Policy Analysis: Philosophical Points of Reference

settled on by the social order. These determine how they should live and encour-age the habit of self-control, so enabling them to be well-ordered selves (Plato,c380s/1952), which means that people’s behavior can be manipulated by argumentsemphasizing duty and obligation. Who a person becomes, then, is contingent uponto whom that person is bound.

Justified Knowledge—To naturalist-structuralists, any statement that cannot be veri-fied by means of abstract thinking and the scientific method by experts would berelegated to the status of expressions of emotion or belief and their validity wouldbe denied.

Rationality Mode—Naturalist-structuralists would prefer to interpret information bythe application of a functionally analytical, albeit a bounded, rationality (Simon1957, 1960, 1976). They would consider the intellect (reason and rational calcula-tion) to be rightly constrained by affects (values and beliefs) imposed by the socialorder. They would prefer to presume a high degree of certainty of both problem-relevant knowledge and agreement on ends. They would prefer to comply withobjectives and value standards settled on by the social order, thereby avoidingpotentially destabilizing conflict over contested or contestable objectives or values.They would also prefer to define problems in a highly structured (reductionist)way.

Decision Analysis—Decision analysis for naturalist-structuralists would involvesearching for satisfactory solutions (Simon 1957, 1960, 1976) that would produceacceptable alternative opportunities for improvement (perhaps conceived asachieving satisfactory balances between the costs involved and the benefits pro-duced, effectiveness achieved or efficiency gains attained) given available resources,which would then be prioritized by their importance, urgency, and values contestability.

Power and Compliance—Naturalist-structuralists would prefer to exercise French andRaven’s (1959) legitimate or expert power, Boulder’s (1990) threat power, orHales’s (2001) knowledge power, and so are inclined toward rules that “ask,command, demand, permit [and] caution” (directive rules [Onuf 1989, p. 86]). Thisis expected to solicit voluntary compliance on the basis of a cognitive commitmentderived from rational calculations made in the context of structural processes (suchas rules and procedures prescribed by those in or with authority). Voluntary com-pliance would be expected—as a matter of a habit—grounded in a sense of obli-gation to obey a sanctioned command.

Salient Analytical Risks—Naturalist-structuralists face the salient analytical risks ofbeing unable to understand either the nature and causation of problems thatcannot be analyzed and explained by the application of naturalist methods, or whytheir solutions, which presume a structuralist ontology, are unable to secure thebehavioral responses required to address policy problems or issues they have iden-tified from free individuals who are not motivated by arguments grounded in dutyand obligation.

The Conduct of Policy Analysis 567

Page 10: The Conduct of Policy Analysis: Philosophical Points of Reference

The Hermeneutic-Structuralist Perspective

Policy analysts predisposed to a hermeneutic-structuralist philosophical stancewould consider the social world to be a subjective social reality, knowable only asit is socially constructed, which they would characterize as a collection of commu-nities of interest with which individuals voluntarily engage. They would presumethat people conduct their affairs in a social order in which no one has a preor-dained place, but everyone belongs and is committed: “we decide what we will think.”

Ethics and Morality—Hermeneutic-structuralists would be inclined toward moral relativism (that there is diversity of moral judgments across time, societies and, individuals [descriptive relativism] and that there is no single true or most-justifiedmorality [meta-ethical relativism]; Foot, 1978). Morality must, thus, draw upon sharedmoral experience and consciousness in a particular milieu, where the way peopleshould conduct their social affairs is cultivated as a set of jointly affirmed socialnorms (rule-governed customs and practices) that constitute the rules by whichpeople should conduct themselves. These social norms reflect language gamesplayed out over time in a way that enables those engaged to make sense of theirmorality, thereby enhancing their moral understanding (Johnston, 1989). As theyplace high value on personal relationships, they would ground their morality onthe principle of character ethics (virtue ethics), premised on a moral act being onein which a good action must be accompanied by good intentions, and the rightemotions and feelings. This makes human qualities more important than abstractethical principles, thereby focusing attention on the Aristotelian qualities of mind(intellectual virtues) and character (moral virtues), which makes such an action intrin-sically virtuous (Aristotle, 334–322 BC/1996).

Human Nature—The hermeneutic-structuralists’ metaphysical explanation of whatit is to be a human being would depend on the sources from which they derivetheir “virtuous qualities.” Individuals who look to, and believe in, received wisdomare forever striving for the right knowledge, the right speech and the right conductthat will lead to salvation. To those who prefer to socially construct their model ofthe human being, they would look to philosophical traditions that emphasize coop-eration: the communitarian tradition, with its underpinning notions of justice, equal-ity, and fairness, whereby individuals have identities that are intimately related togroup membership and are forever striving for robust common ends, involvingbenevolence (Hutcheson) and a just and legitimate civil order (Rousseau); and thephilosophical communism tradition, whose underpinning notions embrace communalcooperation, perhaps as a means of achieving higher ideal, whereby individuals areessentially creative, altruistic, and gregarious beings (Marx). Human nature, then,is circumstantial, a product of people’s social formations, which means that people’sbehavior can be manipulated by arguments grounded on jointly affirmed socialnorms. Who a person becomes, then, is contingent upon who that person knows.

Justified Knowledge—Hermeneutic-structuralists would deny that informationderived by means of inductive or deductive reasoning is, ipso facto, the truth. To

568 John Dixon and Rhys Dogan

Page 11: The Conduct of Policy Analysis: Philosophical Points of Reference

them, the veracity of the knowledge claims can only be established thorough discussion that enables the collective construction of appropriate mutual under-standings. Good argument and the validity of normative judgments thus becomesthe final authority (Bakhtin, 1981; Foucault, 1978; Gergen & Thatchenkery, 1998).They would, therefore, treat knowledge as if it were imperfect, but in a holistic way.

Rationality Mode—Hermeneutic-structuralists would prefer to interpret informationby the application of critical rationality, involving the building of group consensusthrough discourse on contestable values and standards. This would involve the useof intersubjective communications to construct mutual understandings, as the basisfor reasoning and for reaching an agreed understanding (de Haven-Smith, 1988),by means of processes in which all committed actors are empowered and enabledto make and question arguments. They would prefer problems to be defined in amoderately structured way, presuming a moderate degree of certainty of problem-relevant knowledge, but they would anticipate intense disagreement about ends,with the inevitable conflict shaped by socially constructed notions of justice, equal-ity, and fairness. They would, thus, place stress on the valuative problem dimen-sion—the desirability of the impact of means on ends.

Decision Analysis—Decision analysis for hermeneutic-structuralists would involvecontinually striving to unearth the collectively determined sensible and practicablegood, achieved by a group consensus on the desirable outcomes with respect towho gets what, when and how, which would, then, constrain the search for alter-native solutions.

Power and Compliance—Hermeneutic-structuralists would prefer to exerciseBoulder’s (1990) integrative power, French and Raven’s (1959) personal or refer-ent power, or Hales’s (2001) normative power. They thus prefer rules that statebeliefs intended to be accepted (instructive rules (Onuf, 1989, p. 85; but see also p. 120). Voluntary compliance would be expected, on the basis of moral commit-ment (Etzioni, 1961).

Salient Analytical Risks—Hermeneutic-structuralists face two salient analytical risks.The first is that hermeneutic knowledge is subject to severe relativism, dynamic, andthus open to constant revision, which makes explanation contingent and predic-tion problematic. The second is that solutions to policy problems and issues thatpresume structuralist ontology are unable to secure the required behavioralresponses from individuals who are not morally commitment to the informingsocial norms.

Hermeneutic-Agency Perspective

Policy analysts predisposed to a hermeneutic-agency philosophical stance(Goffman, 1959/1990) would deny the possibility of an objective social reality and,therefore, the predictability of social action. They would presume themselves to beincapable of describing, analyzing and understanding—let alone changing—social

The Conduct of Policy Analysis 569

Page 12: The Conduct of Policy Analysis: Philosophical Points of Reference

reality with any degree of certainty. They would consider that no experience canbe fully shared by two individuals. All that is knowable, then, is what is in a person’sown field of contemporaneous consciousness, which cannot be escaped: “I am myworld” (Wittgenstein, (1922/1961, 5.63), “the world is my idea” (Schopenhauer,1818/1844/1969, p. 1). In this world, individuals simply exist. It is up to them notonly to decide their own fate, for which they alone are responsible, but also todefine their own identity, or essential characteristics, which they do in the courseof living out their lives in the most authentic and fulfilling way possible (Heidegger, 1927/1967; Nietzsche, 1986/1966; Sartre, 1946/1973. The socialcontext for hermeneutic-agents is one where everyone has a place but no onebelongs, and where everyone is continually striving to discern the true reality as itis, and as it becomes (manifests), which includes all that which an individual believesto be real: “I guess what I must think.”

Ethics and Morality—Hermeneutic-agents, as moral skeptics, would readily deny thatveracity of any moral knowledge, either because it is not possible to obtain the evi-dence to justify any moral beliefs, or because moral beliefs can never be anythingmore than just matters of personal taste or preference, perhaps the product of intu-ition (Dummett, 1977; Hyting, 1956/1971). They would adhere to the propositionthat moral statements derived from moral norms (moral judgments) express moralemotions (emotive responses to other people, events, situations, points of view, andprinciples [emotivism]) (Hägerström, 1964; Stevenson, 1963), which reveal moral sentiments (those feelings that are central to moral agency) (Gibbard, 1990). Moral conduct is, thus, grounded in the existential ethical proposition that peoplehave to confront ethical dilemmas and make their own ethical choices and toexpress their own moral judgments. They can either face up to their moral respon-sibilities, as they see them, for self and others, or avoid them; the choices they make in specific situations, however paradoxical, ambiguous, and absurd, deter-mine what they make of themselves (Olafson, 1967). This means that no one canbe held responsible to others, morally or otherwise, for their actions (Cicero, 44BC/1971).

Human Nature—The hermeneutic-agents’ metaphysical explanation of what it is tobe a human being draw upon a number of philosophical traditions: existentialism(individuals are essentially unique beings free to choose who they are and how theywish to behave) (Kierkegaard and Sartre); nihilism (individuals live a life free of anyresponsibilities, obligations or duties grounded in any belief systems or moral orreligious principles) (Chernyshevskii, Bakunin, Herzen, Nietzsche, and Pisarev);and anarchism (individuals are born free and nobody is fit to rule over them)(Godwin and Proudhon). Their perception of human nature can be characterizedas: “Man is nothing else but that which he makes of himself ” (Sartre, ([1946] 1973,p. 28), which means that people’s behavior is not instrumentally manipulatable byothers. Who a person becomes, then, is contingent upon who that person wishesto be.

Justified Knowledge—Hermeneutic-agents would not give credence to any knowl-edge purporting to be the truth. They would argue that the canons of rationality,

570 John Dixon and Rhys Dogan

Page 13: The Conduct of Policy Analysis: Philosophical Points of Reference

validity, truth, and efficiency are simply irrelevant. They would, instead, prefer todeal with what is real by discerning the true reality as it manifests to them.

Rationality Mode—Hermeneutic-agents would be predisposed to nonrationality(Elster, 1985; Popkin, 1979; Portes, 1972) as a mode of reasoning. This means thatpolicy problems and issues are situated in a risky, unstable, and unknowable envi-ronment. They would subscribe to the belief that what is important is to make senseof reality, assuming that human behavior is unpredictable. They would, thus, sub-scribe to Weick’s (1995) perception that sense making involves rolling or serialhindsight and is driven by plausibility rather than accuracy. They would preferproblems to have neither a structure imposed upon them nor their causation presumed to be understandable. All possible solutions should be kept open to maximize opportunities for at least preventing the worst outcome or minimizingdamage, as coping strategies.

Decision Analysis—Decision analysis for hermeneutic-agents would involve randomsearch behavior, inspiration, and risk minimization, with any decision being theproduct of garbage-can-like decision processes (March & Olsen, 1976). They wouldprefer to presume no certainties and to avoid contesting problem knowledge, ends, or means by reference to their unknowability. Thus, surprise dominates.Intelligence cannot improve ignorance, and goals and values are luxuries. Onlynondecisions or incremental decisions make any sense.

Power and Compliance—Hermeneutic-agents would expect the powers that be toexercise Boulder’s (1990) destructive threat power or French and Raven’s (1959)physical or coercive power over them. They would, however, deny any voluntarycompliance obligation, which means they comply only under coercion, so givingrise to alienative compliance (Etzioni, 1961).

Salient Analytical Risk—Hermeneutic-agents face the salient analytical risk that theirdenial of an objective social reality leads to an exaggeration of the unpredictabil-ity of social action. Even if all knowledge is based on personal experience, an individual’s taken-for-granted “stock of knowledge” (natural attitude [Husserl,1931/1960]) itself is based on the expectation of reciprocity which, while never complete, may be so near as to provide high probability of quasi-prediction and,by implication, the possibility of quasi-structuralist causation.

The Underclass: A Demarcation of Contending Meanings, Causations and Solutions

The concept of “underclass,” as a residuum or marginalized social group (see, forexample, Cottingham, 1982; Dean & Taylor-Gooby, 1992; Hyde & Armstrong,1995; Lister, 1990; MacDonald, 1997; Macnicol, 1994; Mann, 1994; Murray, 1996)evokes a multiplicity of attitudes and beliefs about its

The features of the underclass vary enormously . . . the disparate observers who claim to be able to witness an underclass are incapable of agreeing on what it is they have witnessed. (Mann, 1994, pp. 79–80)

The Conduct of Policy Analysis 571

Page 14: The Conduct of Policy Analysis: Philosophical Points of Reference

This occurs because observers disagree on the meaning, even the existence, of the“underclass,” not to mention its causes and solutions, because they hold incom-patible contentions about what constitutes a fact in the social world, what has causalcapacity, and how human nature can best be characterized.

Naturalist-Agency Perspective

The “underclass” to those of this philosophical disposition is objectively measur-able (against, for example, low levels of skills, severe income depravation, depen-dence on welfare, poor employment potential, limited education, health problems)and is a product of an individual’s self-interested calculation of material net bene-fits given their particular circumstances, which has its foundations in the idea ofmoral hazard. This leads, then, to their failure to acquire the sense of status andidentity that flows from achieving material success. Their solution to the “under-class problem”—the problem of free-riding individuals sustaining a dependencylifestyle at the taxpayers expense—is to curtail generous welfare programs thatsimply encourage weak-willed individuals to remain noncompetitive, that provideno financial incentive to compete, and that make no attempt to change personalvalues; and to substitute programs that facilitate their engagement in the labormarket.

Naturalist-Structuralist Perspective

The “underclass” to those of this philosophical disposition is also objectively measurable, but it is a product of people loosing their sense of social identity and,therefore, their acceptance of binding social obligations and responsibilities. Theirsolution to the “underclass problem”—a weakening of the compliant hierarchicalsocial order—is to initiate controlling forces (perhaps in the form of tougher lawsand more diligent policing) that reinforce more orderly and conformist behaviorin those considered to be deviant, which they would consider to be naturally accept-able to everybody. By these means, an end would be brought to the normlessnessand unpredictability of the relatively random world in which members of the“underclass” live.

Hermeneutic-Structuralism Perspective

The “underclass” to those of this philosophical disposition is a social construct, andis a product of people being engaged in an alternative form of life that prevents,to their disadvantage, their engagement with mainstream discourse. Their solutionto the “underclass problem”—the disempowerment of a vulnerable and alienatedgroup in society—is to persuade those deemed to be disempowered and alienated,and thus who feel marginalized from social life, that they should engage withempowering structures seeking common ends (perhaps involving benevolence, ajust and legitimate civil order, or communal cooperation). Thus, processes thatbuild social engagement with the disengaged and disempowered are needed, whichwould enhance individual growth and development, thereby enabling them tobecome essentially creative, altruistic, and gregarious beings.

572 John Dixon and Rhys Dogan

Page 15: The Conduct of Policy Analysis: Philosophical Points of Reference

Hermeneutic-Agency Perspective

To those of this philosophical disposition, the “underclass” is self-referential andcannot exist outside the consciousness of those who identify with its defining attrib-utes and characteristics. The self-defining members of the underclass condemnthemselves to a life that is inauthentic and unfulfilling. Their solution to the“underclass problem”—the inability of some people to make something of them-selves—is to help those who consider themselves to be members of the underclassto embrace their existential isolation, to accept the vicissitudes that flow from it,and to proceed, regardless, toward the construction of their own futures, therebyfulfilling their unique potential. This can be achieved by creating convincing oppor-tunities for people to contribute to the design of public policy programs targetingthem.

The Philosophical Conditions for Coherent Policy Analysis

The array of methodological families summarized in Figure 1 poses challengingquestions about how policy analysis should be conducted. In the face of incom-patible epistemological and ontological contentions, the contemporary philosophyof the social sciences offers a way forward, through the diverse philosophical andmethodological work of Archer, Bhaskar, Bourdieu, and Giddens. They have her-alded a very clear attempt to reconcile the epistemological limitations of naturalismand hermeneutics, resulting in the transcendental realism synthesis (Bhaskar, 1998), andthe ontological inadequacies of structure and agency, resulting in the poststructurationsynthesis (Archer, 1995; Bourdieu, 1998; Giddens, 1984, 1993).

Transcendental realism, the epistemological synthesis, is concerned with the natureof factual description of the real world and offers a process by which its causalmechanisms can be identified. It makes two fundamental claims. The first is thatthe real world operates at three levels: the empirical (the perceived nature of eventsor processes open to the observer); the actual (events or processes as they actuallyoccur); and the real or deep (the underlying mechanisms or imperatives that causethese events or processes) (Baert, 1998, p. 191). Knowledge of the real world, then,rests on unreliable empirical perceptions of the actual world, which is itself oneremove from any deep explanations of it. Bhaskar (1998, p. 11) thus draws a dis-tinction between transitive objects or phenomena (objects or phenomena as theyare experienced) and intransitive objects or phenomena (objects or phenomena asthey are in reality). The second claim is that explanation of the real world involvesa cumulative process of hermeneutic-based imaginative model building, wherebytransitive knowledge is used to postulate hypothetical causal mechanisms that, ifthey exist, would explain any intransitive phenomenon under investigation. Tran-scendental realism does not overcome the uncertainties identified by the earliernaturalists, for problems of induction and the theory-laden nature of observationremain (Popper, 1935/1959/2000). However, it does embrace them at the ontolog-ical level, and it does adopt more sophisticated criteria for reality, which are freefrom the constraints of strict falsificationism. It also offers a potential reconciliationof the hermeneutic aspects of scientific discovery identified by Kuhn (1970) withan empirical-based approach to inference to the best explanation, that is the

The Conduct of Policy Analysis 573

Page 16: The Conduct of Policy Analysis: Philosophical Points of Reference

process of choosing the hypothesis or theory that best explains available data(Wendt, 1999).

The poststructuration ontological synthesis is an attempt to adjudicate the ontologi-cal tensions between structure and agency. In contention is whether agency and struc-ture are interdependent, in a duality relationship as asserted by Giddens: “Thereflexive capacities of the human actor are characteristically involved in a contin-uous manner with the flow of day-to-day conduct in the contexts of social activity”(1984, p. xxiii), or interdependent but different and thus distinguishable (in ananalytically dualist; Bhaskar, 1975) or morphogenetic relationship (Archer, 1995),which means that, with time and power, social structure is both a cause and a con-sequence of agency (Parker, 2000).

The combination of transcendental realism and poststructuration suggest a fifthmethodological position. This is one that presumes a world in which events orprocesses are knowable, the nature of which, however, can be only unreliably andcontestedly perceived by an observer, and in which structure and agency only haveproperties that are manifest in, and reproduced or transformed through, socialpractice. The knowledge so gained can be used to generate hypothetical causalexplanations for the observed events or processes, for which empirical corrobora-tion can be sought. The discovery of an intransitive generative mechanismbecomes, itself, a new phenomenon that needs to be explained. Progressively,deeper levels of explanation of the social world are thereby generated by thismethodology. When applied to the analysis of policy, this fifth methodological posi-tion progressively facilitates deeper levels of understanding of the specific policy-relevant events and processes, permits more subtle explanations of problem orissue causation, and facilitates the enhancement of policy learning through thereflexive capacities of those it empowers.

In seeking to understand the causes of policy problems or issues, transcendentalrealists would be skeptical of any empirical generalizations about their causationderived from naturalist methods, which they would treat only as preliminaryworking hypotheses. They would search for a deep understanding of the under-lying causation mechanism or imperatives. This would require them to engage withothers concerned about or engaged with the policy problem or issue in acts ofreflexive interpretation of the ideational and material dimensions of the problems,so as to ensure that they have an appropriate contextualization of meaning, whichwould involve the application of hermeneutic methods that would enable them toidentify perspective reciprocities that result from acts of ideation. This cumulativeprocess of hermeneutic-based imaginative model building involves transitive knowl-edge being used to postulate hypothetical causal mechanisms that, if they can beempirically demonstrated to exist, would explain the policy problem or issue cau-sation. This would involve a search for empirical corroboration. If such confirma-tion were possible then a new intransitive generative mechanism would have beendiscovered, which would, in turn, become a new phenomenon to be explained.Transcendental realism thus leads progressively to deeper levels of explanation of the policy problem or issue causation, thereby permitting more subtle causal explanations.

In seeking to identify how best to deal with policy problems or issues for whichsubtle explanation has been found and agreed upon, poststructurationists would

574 John Dixon and Rhys Dogan

Page 17: The Conduct of Policy Analysis: Philosophical Points of Reference

accept that those concerned about or engaged with the policy problem or issuehave the necessary reflexive capacities to solve them, but those capacities can onlybe actualized, so becoming meaningful human action, when all participants areempowered and enabled to draw upon the structural properties of the policysystem. This reflexive capacity is the embodied understanding they gain of theideational and material dimensions of a policy problem or issue by engaging withpolicy practice, thereby enabling them to learn by trail and error and from the mis-takes made by others, so as to determine the relevance of general principles (suchas rules, recipes, formal procedures, and judgmental criteria). By this means theyare able to garner the understanding needed to solve policy problems or issues asthey conduct their affairs within the policy system. The resultant policy practice,mobilized as it is in a continuous manner with the flow of day-to-day conduct, will,in turn, transform the enabling structural properties of the policy system. Thiscreates the potential for further policy problems or issues, so necessitating theprospect of further learning as the search for problem or issue explanation andsolution continues.

The acceptance of this synthetic philosophical stance, however, generates seriousepistemological and ontological challenges for those engaged in policy analysis.

Implications for the Intellectual Integrity of Policy Analysis

The proposed fifth methodology requires policy analysts to be philosophicallyreflective, and thus able not only to identify their own and others epistemologicaland ontological predispositions, but also to understand and accept the strengthsand weaknesses of the contending methodologies for their performance as policyanalysts. In essence, this requires them to embrace the following propositions.

First, adept policy analysts would be epistemologically and ontologically sophis-ticated enough to accept that what constitutes “good” policy analysis is an essen-tially contested concept, clarifiable through constructive discourse. Thus, theywould actively seek insights into what might work in particular policy situations byengaging with those who hold different philosophical dispositions. They thereforewould see such constructive discourse as normal, even if it has the propensity tocreate conflict, and, most certainly, as necessary, in order to create creative oppor-tunities for policy analysts to engage with those holding contending philosophicalperspectives to find solutions to policy problems and issues.

Second, adept policy analysts would be skeptical of any empirical generalizationsabout the causation and consequences of, and solutions to, policy problems andissues. These they would treat only as preliminary working hypotheses. They wouldthus seek to deepen their understanding by engaging in acts of ideation with thosewho hold different philosophical dispositions, which would allow the perspectivesreciprocity needed for a reflexive interpretation to emerge that would ensure anappropriate contextualization of meaning.

Third, adept policy analysts would learn how to comprehend and evaluate theintended meaning of the contending arguments based on a diversity of epistemo-logical and ontological perspectives. They would settle in their own minds com-peting epistemological and ontological truth-claims with consistency and withoutrecourse to intentional activities and motivated processes that enable self-deception

The Conduct of Policy Analysis 575

Page 18: The Conduct of Policy Analysis: Philosophical Points of Reference

or self-delusion. They would thereby confront unpleasant truths or issues ratherthan resort to the mental states of ignorance, false belief, unwarranted attitudes,and inappropriate emotions (Haight, 1980). They would accept that the best policyoutcomes that can be expected from constructive discourses are sets of achievablepolicy aspirations, implementable strategies, and a tolerable level of policy conflict.They would see view good policy analysis as an iterative process that involves learn-ing-by-doing and learning-from-experience about what is the right thing to do andhow to do things right.

Conclusion

In a world characterized by profound diversity of opinion grounded in equally pro-found philosophical differences, the interrogation of the social world requires aseries of conceptual and analytical tools. These relate not only to epistemology andontology, but also to methodology, and thus theory and method of inquiry. Fur-thermore, these various tools are logically interconnected: epistemology shapesontology, epistemology and ontology together shape methodology, and methodol-ogy shapes both theory and method. Adept policy analysts must be critically reflec-tive before they seek to describe, explain, understand, judge, and address policyproblems and issues by drawing upon theories and methods grounded in only oneof these contending (fundamentally flawed) methodological families. The broadconclusion drawn, then, is that policy analysts need:

• to avoid epistemological and ontological arrogance;

• to seek out and engage with those who disagree with their philosophical dispositions;

• to treat all truth-claims skeptically, accepting that there are multiple standardsby which they could be justified, particularly if they come from any ascendantepistemic community (whether grounded in naturalism or hermeneutics); and

• to settle competing epistemological and ontological asseverations with consis-tency and without recourse to the self-deception or self-delusion that permitsthem to avoid unpleasant truths about their informing intellectual discipline.

About the Authors

John Dixon is Professor of Public Management at the University of Plymouth in the UnitedKingdom, where he is the Director of the Governance Centre. He has published extensivelyin the filed of public and social policy and management. His latest book is Responses to Governance: Governing Corporations, Societies and the World (Praeger, 2003).

Rhys Dogan is Principal Lecturer in Politics at the University of Plymouth in the UnitedKingdom. He has published in the field of European integration theory, as well as the philosophy of the social sciences.

References

Archer, M. S. (1995). Realist social theory: The morphogenetic approach. Cambridge: Cambridge University Press.Aristotle. (1996). Nichomachean ethics (T. Griffith, Trans.). Ware: Wordsworth Classics. (Original work published

334–322 BC).

576 John Dixon and Rhys Dogan

Page 19: The Conduct of Policy Analysis: Philosophical Points of Reference

Baert, P. (1998). Social theory in the twentieth century. Cambridge: Polity Press.Bakhtin, M. (1981). The dialogic imagination. Austin: University of Texas Press.Bentham, J. (1970). An introduction to the principles of morals and legislation. London: Athlone. (Original work

published 1789).Bhaskar, R. (1975). A realist theory of science. Leeds, Yorkshire, United Kingdom: Leeds Books.Bhaskar, R. (1998). The possibility of naturalism: A philosophical critique of the contemporary human sciences (3rd ed.).

London and New York: Routledge.Blumer, H. (1969). Symbolism interactionism: Perspective and method. New York: Prentice Hall.Boulder, K. (1990). Three faces of power. Newbury Park, CA: Sage.Bourdieu, P. (1998). Practical reason. Cambridge: Polity Press.Brink, D. (1989). Moral realism and the foundations of ethics. Cambridge: Cambridge University Press.Christman, J. (Ed.). (1989). The inner citadel: Essays on individual autonomy. Oxford: Oxford University Press.Cicero. (1971). De fato [On fate] (tr Sharples, R. W.). Warminister, Wiltshire, United Kingdom: Aris & Phillips.

(Original work published 44 BC).Cottingham, C. (1982). Race, poverty and the urban underclass. Lexington, DC: Heath.de Haven-Smith, L. (1988). Philosophical critiques of policy analysis. Gainesville: University of Florida Press.Dean, H., & Taylor-Gooby, P. (1992). Dependency culture: The explosion of myth. Hemel Hempsted, Hertsford,

United Kingdom: Harvester Wheatsheaf.Dewey, J. (1985). The later works 1925–1953 (Vol. 7) Ethics. Carbondale: Southern Illinois University Press.

(Original work published 1932.)Dixon, J. (2003). Responses to governance: The governing of corporations, societies and the world. Westport, CT:

Praeger.Dixon, J., & Dogan, R. (2002). Hierarchies, networks and markets: Responses to societal governance failures.

Administrative Theory & Praxis, 24(1), 175–196.———. (2003a). A philosophical analysis of management: Improving praxis. Journal of Management Develop-

ment, 22(6), 458–482.———. (2003b). Analysing global governance failure: A philosophical framework. Journal of Comparative Policy

Analysis, 5(1), 207–224.———. (2003c). Corporate decision making: Contending perspectives and their governance implications.

Corporate Governance: The International Journal of Business in Society, 3(1), 39–57.———. (2003d). Towards constructive corporate governance: From “certainties” to a plurality principle. Reason

in Practice: The Journal of the Philosophy of Management, 2(3), 51–72.Dummett, M. A. E. (1977). Elements of intuitionism. Oxford: Clarendon.Dumont, L. (1970). Homo Hierarchus. Chicago: University of Chicago Press.Dworkin, G. (1988). The theory and practice of autonomy. Cambridge: Cambridge University Press.Elster, J. (1985). Sour grapes: Studies in the subversion of rationality. Cambridge: Cambridge University Press.Etzioni, A. (1961). A comparative analysis of complex organizations. New York: Free Press.Fletcher, J. (1966). Situational ethics. Philadelphia, PA: Westminster Press.Foot, P. (1978). Moral relativism. Lawrence: University of Kansas Press.Foucault, M. (1978). The history of sexuality (Vol. 1): An introduction (R. Hurley, Trans.). New York: Pantheon.French, J. R. P., & Raven, B. (1959). The bases of social power. In D. Cartwright (Ed.). Studies in social power

(pp. 150–167). Ann Arbor: University of Michigan Press.Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.Gergen, K. J., & Thatchenkery, T. J. (1998). Organizational science in postmodern context. In R. C. H. Chia

(Ed.). In the realm of organization: Essays for Robert Cooper (pp. 15–42). London: Routledge.Gibbard, A. (1990). Wise choice, apt feelings: A theory of normative judgment. Cambridge, MA: Harvard University

Press.Giddens, A. (1984). The constitution of society. Cambridge: Polity Press.———. (1993). New rules of sociological method: A positive critique of interpretative sociologies (2nd ed.). Stanford:

Stanford University Press.Goffman, E. (1990). The presentation of self in everyday life. Harmondsworth: Penguin. (Original work published

1959.)Gouinlock, J. (1972). John Dewey’s philosophy of value. New York: Humanities Press.Hägerström, A. (1964). Philosophy and religion (R. T. Sandin, Trans.). London: Allen and UnwinHaight, M. R. (1980). A study of self-deception. London: Humanities Press.Hales, C. (2001). Managing through organization: The management process, forms of organisation and the work of

managers. London: Business Press.Hampton, J. (1986). Hobbes and the social contract tradition. Cambridge: Cambridge University Press.Heidegger, M. (1967). Being and time (J. Macquarrie & E. Robinson, Trans.). Oxford: Basil Blackwell.

(Original work published 1927.)Hempel, C. G. (1966). Philosophy of natural science. Englewood Cliffs, NJ: Prentice Hall.Hollis, M. (1994). The philosophy of social science: An introduction. Cambridge: Cambridge University Press.

The Conduct of Policy Analysis 577

Page 20: The Conduct of Policy Analysis: Philosophical Points of Reference

Hollis, M., & Lukes, S. (Eds.). 1982. Rationality and relativism. Cambridge, MA: MIT Press.Husserl, E. (1960). Cartesian meditations (D. Cairns, Trans.). The Hague, Netherlands: M. Nijhoff. (Original

work published 1931.)Hyde, M., & Armstrong, E. (1995). Underclass or underdogs? Britain’s poor in the 1990’s. Public Policy Review,

3(4), 44–46.Hyting, A. (1971). Intuition: An introduction (3rd ed.) Amsterdam: North Holland. (Original work published

1956.)Johnston, P. (1989). Wittgenstein and moral philosophy. London: Routledge.Kant, I. (1996). Critique of practical reason (T. K. Abbott, Trans.). New York: Prometheus. (Original work

published 1788.)———. (1998). Critique of pure reason (P. Gruyer & A. W. Wood, Eds. and Trans.). Cambridge: Cambridge

University Press. (Original work published 1781.)———. (2002). Groundwork of the metaphysics of morals (A. W. Wood, Trans.). New Haven, CT: Yale University

Press. (Original work published 1785.)Kuhn, T. S. (1970). The structure of scientific revolutions (2nd ed.). Chicago: University of Chicago Press.Lister, R. (1990). The exclusive society: Citizenship and the poor. London: Child Poverty Action Group.MacDonald, R. (1997). Youth, the Underclass and Social Exclusion. London: Routledge.Macnicol, J. (1994). The underclass debate: Views from history. Journal of Social Policy, 23, 162–190.Mann, K. (1994). Watching the defectives: Observers of the underclass in the USA, Britain and Australia.

Critical Social Policy, 4, 79–99.March, J. G., & Olsen, J. P. (1976). Ambiguity and change in organizations. Bergen, Norway: Universitets Forlaget.Marks, J. (Ed.). (1986). The way of desire. Chicago: Precedent.Meyers, G. E. (1986). William James: His life and thought. New Haven, CT: Yale University Press.Mill, J. S. (1968). Utilitarianism, liberty and representative government, London: Everyman. (Original work pub-

lished 1863.)Moser, P. K. (1989). Knowledge and evidence. Cambridge: Cambridge University Press.Muirhead, J. H. (1932). Rule and end in morals. Oxford: Oxford University Press.Murray, C. (1996). The emerging British underclass. London: IEA Health and Welfare Unit.Nietzsche, F. (1966). Beyond good and evil (W. Kaufmann, Trans.). New York: Random House. (Original work

published 1886.)Nightingale, D. J., & Cromby, J. (1999). Social constructionist psychology: A critical analysis of theory and practice.

Buckingham, Bucks., United Kingdom: Open University Press.Olafson, F. (1967). Principles and persons: An ethical interpretation of existentialism. Baltimore, MD: Johns Hopkins

University Press.O’Neill, O. (1989). Constructions of reason: Explorations of Kant’s Practical Philosophy. Cambridge: Cambridge

University Press.Onuf, N. G. (1989). World of our making: Rules and rule in social theory and international relations. Columbia:

University of South Carolina Press.Parker, J. (2000). Structuration. Buckingham: Open University.Pitt, J. (1989). Theories of explanation. Oxford: Oxford University Press.Plato. (1952). Phaedras (R. Hackforth, Trans.). Cambridge: Cambridge University Press. (Original work

published c380s BC.)Popkin, S. (1979). The rational peasant: The political economy of rural society in Vietnam. Berkeley: University of

California Press.Popper, K. (2000). The logic of scientific discovery (2nd ed.). London: Routledge. (Original work published

1935/1959.)Portes, A. (1972). Rationality in the slum: An essay on interpretive sociology. Comparative Studies in Sociology

and History, 14, 668–686.Rawls, J. A. (1971). A theory of justice. Oxford: Oxford University Press.Sartre, J. P. (1973). Existentialism and humanism (P. Mairet, Trans.). London: Methuen. (Original work published

1946.)Scarre, G. (1996). Utilitarianism. London: Routledge.Scheffler, S. (1988). Consequentialism and its critics. Oxford: Oxford University Press.Schopenhauer, A. (1969). The world as will and representation (E. F. J. Payne, Trans.) (2 vols.). New York: Dover.

(Original work published 1818 and 1844).Schutz, A. (1967). The phenomenology of the social world. Evanston, Ill.: Northwestern University Press. (Original

work published 1932.)Simon, H. (1957). Models of man: Social and rational. New York: Wiley.Simon, H. A. (1960). Administrative behavior (2nd ed.). New York: Macmillan.———. (1976). Administrative behavior (3rd ed.). New York: Macmillan.Stevenson, C. L. (1963). Facts and values: Studies in ethical analysis. New Haven, CT: Yale University Press.

578 John Dixon and Rhys Dogan

Page 21: The Conduct of Policy Analysis: Philosophical Points of Reference

Weick, K. E. (1995). Sensemaking in organizations. Newbury Park, CA: Sage.Wendt, A. E. (1999). Social theory of international politics. Cambridge: Cambridge University Press.Williams, M., & May, T. (1996). Introduction to the philosophy of social research. London: UCL Press.Winch, P. (1990). The idea of social science and its relation to philosophy (2nd ed.). London: Routledge.Wittgenstein, L. (1961). Tractatus Logico-Philosophicus (D. F. Peters & B. F. McGuiness, Trans.). London:

Routledge. (Original work published 1922.)Woodfield, A. (1976). Teleology. Cambridge: Cambridge University Press.

The Conduct of Policy Analysis 579