Chapter 13. Rational Choice in Public Policy: The Theory in Critical Perspective

174



Handbook of Public Policy Analysis



THE BUILDING BLOCKS OF RATIONAL CHOICE THEORY

Rational choice theory is best characterized as a school or an approach to understanding the dynamics of public policy; it is a “family of theories” rather than a single theory (Green and Shapiro

1994, 28). The likes of game theoretical accounts of strategic interactions (Axelrod 1984), social

choice critiques of the impossibility of a democratic aggregation of interests (Arrow 1951), and

public choice applications of economics to politics (Niskanen 1971; Olson 1982) all sit shoulder

to shoulder under the auspices of rational choice. In this opening section, we identify a set of common building blocks that arguably provide the foundations that hold such mainstream rationalist

approaches together; foundations that have however been stretched or applied over time with varying

levels of consistency as we shall discover below. Our first common building block is the assumption

of rationality, which posits that individuals in the policy process are rational, and that the behavior

of agents is best explained by imputing rationality on to them (Dowding and King 1995, 1). We

then go on to examine further heroic qualities associated with individuals in the form of consistent

and rank-ordered preferences before addressing the premises of methodological individualism and

deductive reasoning.



THE ASSUMPTION OF RATIONALITY

Within rational choice accounts of public policy, individuals are characterized as instrumental actors,

pursing courses of actions “not for their own sake, but only insofar as they secure desired, typically

private ends” (Chong 1996, 39). In this instance, rationality is entwined with utility-maximization.

Rational individuals are those individuals who, when faced with distinct courses of action or policy

options, choose the feasible course of action, which is most likely to maximise their own utility.

Rationality thus emerges from an actor’s capacity to calculate and attach costs and benefits to available policy options, and to select the course of action that best maximises her own utility (Dunleavy

1991, 3; Majone 1989, 13; Elster 1986, 4). Policy actors are constructed as egotistical, self-regarding

instrumental actors, “choosing how to act on the basis of the consequences for their personal welfare

(or that of their immediate family)” (Dunleavy 1991, 3). As such, mainstream rational choice lends

itself towards explanations of policy outcomes grounded in the goal-oriented action of individuals

(MacDonald 2003, 552) where the desires, beliefs, and preferences of individual actors are identified

as the causes of their actions (Green and Shapiro, 1994: 20). Of course, such rational behavior by

individuals will not necessarily produce positive collective outcomes. Rational utility-maximizing

individuals will and do deliver collectively unintentional outcomes or socially irrational outcomes

(Levi 1997, 20) as the “tragedy of the commons” demonstrates (Hardin 1969). Rational choice thus

explains the irrational collective outcomes of the individual actions of rational actors.



RANK-ORDERED AND CONSISTENT PREFERENCES

As we suggested above, this assumption of rationality hinges upon a number of heroic qualities

attached to individual actors, not least that rational actors possess, even when faced with complex

situations, the capacity (in terms of information, time, and objectivity) to choose the optimum course

of action open to them (Ward 2002, 68). Importantly, individual rational actors must, if the assumption of rationality is to hold, possess preferences or wants that are ranked-ordered, and consistent

in that they are both transitive and stable over time. First, the rank-ordering of preferences asserts

that individuals are able to establish a hierarchy of wants or preferences; preferences of course

have to be comparable. Second, the condition of transitivity establishes a particular consistency of



Fisher_DK3638_C013.indd 174



9/19/2006 3:47:41 PM



Rational Choice in Public Policy



175



preferences in that if an individual prefers oranges (x) to pears (y) and pears (y) to apples (z), she

will prefer oranges (x) to apples (z). Under such conditions, preferences are said to be transitive.

Finally, an individual’s preferences are taken to be both stable over time (at least in the short-term),

and to be given. In terms of this latter condition, that which assumes the given-ness of preferences,

mainstream rationalist accounts are simply saying that for the purposes of explanation, they are not

overly concerned with the origins of individuals’ preferences. For Shepsle and Bonchek, rational

choice accounts “take people as we find them,” less concerned with “why they want what they

want” (1997, 17).



METHODOLOGICAL INDIVIDUALISM

Mainstream rational choice builds its explanations of policy outcomes on the actions of individuals,

privileging the role of actors over and beyond that of social structures. Social structures are attributed

no “independent status apart from the individuals who constitute them. Only actors choose, prefer,

believe, learn and so on: society does not act independently of them” (Lichbach 2003, 32–33). This

privileging of actors locates the explanation of macro-level phenomena at the micro-level of the

individual, or at the aggregated behavior of individuals (Van Thiel 2004, 180). Social structures are

thereby conceptualized through the intentional behavior of individuals, interpreted as the aggregate

result of the calculations and utility-maximizing strategies of individual actors, such that institutions

become “instrumental products used by individuals to maximise their respective utilities” (Blyth

2002, 19). As such, rational choice stresses the production of causal mechanisms, underplayed in

structural and functionalist explanations (Chong 1996, 42–43). For, as Laver argues, “the primitive

motivational assumptions” of rational choice “relate to the individual as a self-contained unit of

analysis” (1997, 9).



DEDUCTIVE REASONING

In applying deductive reasoning, rational choice explanations of policy outcomes advance through

first establishing sets of propositions and statements about real world phenomena and the rational

behavior of individuals, and then testing these propositions through comparison with events and

the actual behavior of actors (Laver 1997, 4). Indeed, Laver (1997) makes an emblematic defence

of the value of what he calls the a priori approach of rational choice over and beyond predominant

empirical approaches to policy analysis (for another defense see Ward 2002, 69–70). Such predominant empirical approaches, Laver argues, simply extrapolate generalized propositions about

political behavior from “systematic observations of what real people actually do,” which “in the last

analysis more or less says that the world is as it is because that’s how it is” (1997, 10). Drawing upon

the work of Nozick and Rawls, Laver argues, however, that the very value of deductive reasoning

in rational choice accounts lies in the fact that the explanation of political outcomes derives from

motivational assumptions defined in another realm, that of the individual.

What these building blocks of mainstream rational choice theory offer is a “convenient shortcut which makes possible a naturalist science of the political such as is capable of generating

through a process of deduction, testable and predictive hypotheses” (Hay 2004, 40). The complexity of both individual motivations and the decision-making process is thus necessarily mobilised

out of rational choice accounts in the quest for universal and parsimonious explanation. Issues of

interpersonal variation and personal identity are, for example, neatly sidestepped under the guise of

rational instrumental utility-maximising individuals, with preferences given, stable over time, and

common across individuals (see below). However, rather than acknowledging the contested nature



Fisher_DK3638_C013.indd 175



9/19/2006 3:47:41 PM



176



Handbook of Public Policy Analysis



of such oversimplifications, mainstream rational choice theorists laud such interchangeability, for

it “constitutes a conscious effort to apply standards of scientific explanation in the social sciences”

(Tsebelis 1990, 45).



THE SEARCH FOR EQUILIBRIUM SOLUTIONS

In discussing the building blocks of mainstream rational choice approaches, we have set our analysis

within the conditions of parametric decision making. Under the conditions of parametric decision

making, individuals calculate the costs and benefits of particular courses of action with no risk or

uncertainty over either the outcomes of particular courses of action or over how the actions of others

might impinge upon potential outcomes. Particular courses of action are known by individual actors

to lead to specific outcomes, and the actions of other individuals are fixed in that they do not affect

the relationships between the particular courses of action open to an individual and the outcomes

associated with each of these particular courses of action (Ward 2002, 68). Here we now move on

from the conditions of parametric decision making to examine the conditions of strategic decision

making and the implications of such conditions for our understanding of the motivations of rational

individuals in the policy process. Understanding the conditions of strategic decision making brings

to the fore the search for strategic equilibria and questions earlier assumptions of methodological

individualism. It leads us first to clarify the assumption of rationality, moving from the maximisation of utility to the maximisation of expected utility.

In fact, rather than the simplifying assumptions of parametric decision making, most policy

making will take place under the conditions of strategic decision making. Strategic decision-making assumes that actors operate under conditions of interdependence where the actions of others

will impinge upon the relationship between courses of action and potential outcomes; such that

personal outcomes for individuals will depend upon the collective expectations and action of others.

Individual actors are thus obliged to act strategically, to anticipate what from a range of policy

options other actors will do and equally what these other actors believe she herself will do (Elster

1986, 7). Under such conditions, and given that rational actors cannot guarantee the outcomes of

particular courses of action, or necessarily know the consequences of their actions, individuals will

make calculations over the efficacy of particular policy instruments and the probability of particular

instruments delivering potential outcomes. In other words, rational individuals will maximise the

expected utility attached to policy instruments.

This recognition of strategic independence privileges the conceptualisation of the processes of

decision-making as complex interactive games played out between individual policy actors whereby

the coordination of the actions of individual actors is at a premium. In such complex interactive

games, the primary analytical concern of rational choice theorists is the search for, or identification

of, the necessary conditions for optimum strategic equilibrium solutions. Equilibrium solutions occur when individual actors involved “in a recurring course of action are considered as not having

any incentives to deviate from this course” (Tsebelis 1990, 41). Equilibrium solutions are thus by

definition self-reinforcing. However, there is no single equilibrium solution in any game being played

out by rational actors (as the Folk Theorem demonstrates), bringing to the fore the question of why

particular equilibrium solutions emerge over others and how stable interactions ensure between

policy actors. Indeed, accounting for such patterns of stability challenges assumptions of unstable

patterns of competition inherent within mainstream rational choice accounts of public policy.

This search for strategic equilibria is problematized by the logic of collective action, presented

as the difficulties of overcoming strategic coordination in multiple actor games. Recurring games

do offer actors the opportunity of trial-and-error learning (see evolutionary game theory). However, rational choice theory has tended to couch explanations of stability within the framework of

institutions. Institutions, it is argued, place constraints upon actors, lessen information costs, and



Fisher_DK3638_C013.indd 176



9/19/2006 3:47:41 PM



Rational Choice in Public Policy



177



offer “nests” for focal points which when taken together facilitate coordination and cooperation

between actors and explain the emergence of specific equilibrium solutions over others (North 1990,

Goldstein and Keohane 1993). Indeed, Garrett and Weingast (1993, 173–207) demonstrate how the

1979 Cassis de Dijon ruling of the European Court of Justice established the principle of the mutual

recognition of goods and services that facilitated moves toward the European single market through

its capacity to operate as a focal point to overcome collective action and coordination difficulties

between member states keen to guard their domestic interests and uncertain of the outcomes of

any further moves towards economic competition. Ideas thus operate in this instance “ex post as

auxiliary hypotheses designed to explain disconfirming outcomes” (Blyth 2002, 26).

This search to explain stability fuels questions over how far the assertion of methodological

individualism best characterizes rational choice theory. Rational choice accounts of policy making

specify prevailing institutional rules of the game such that, as Dowding and King (1995, 1) assert,

rational choice theory explains how context informs the preferences of actors, offering explanations of political outcomes “where the reasons for individual action are specified, and the structural

conditions under which those reasons of action have been modelled and explained.” The behavior of

individuals becomes an optimal response both to other actors and the dominant institutional rules and

norms, such that “the prevailing institutions (the rules of the game) determine the behaviour of the

actors, which in turn produces political or social outcomes” (Tsebelis 1990, 40); or, in other words,

“preferences provide the motivation of individual action; institutions provide the context allowing

causal explanation” (Dowding and King 1995, 1). This assertion does not necessarily contest the

fundamental assumption of methodological individualism, which is that macro-social outcomes are

the product of the micro-actions of individuals whose behaviour and identity are not determined by

such structures. However, Ward argues that the existence of multiple equilibria requires actors to

coordinate beliefs through common conjectures and that such patterns of beliefs are intersubjective,

forming “a system that cannot be reduced to the beliefs of any one player considered as a ‘social

atom’” (Ward 2002, 71). Indeed, for Ward, rational choice models, “in ontological terms partly

comprise taken-as-given socio-structural elements,” informing us how decisions are taken within

social structures (Ward, 2002, 71).

Overall, therefore, the conditions of strategic decision making pose a number of internal challenges to rational choice accounts of the policy process, not least the explanation of the emergence

of particular stable patterns of policy making or strategic equilibrium solutions. The predominant

response to such challenges has been the turn toward further incorporation into rational choice of

the role of institutions and ideas. However, moves toward institutions and ideas strain assumptions

of rationality and potentially hamper quests for universal explanation. Indeed, the very acceptance

of multiple equilibria weakens the predictive power of rational choice. With these strains in mind,

we now examine the challenges and limits to rational choice theory.



CHALLENGES TO RATIONAL CHOICE THEORY

The rationality assumption that we have addressed so far parsimonious and universal explanations

founded upon the actions of rational utility-maximizing individuals. However, this very assumption

of rationality is often challenged as the Achilles’ heel of mainstream rational choice theory, labeled

as being reductionist and falsely biased toward economistic arguments, as attributing overoptimistic

heroic qualities to individuals, and as lacking empirical foundation (see, for example, the work of

Green and Shapiro 1994). At their strongest, such challenges summarily dismiss short-term utilitymaximization as historically and culturally specific to capitalist societies, with rational choice theory

characterized as “more an ideological expression of the United States’ interests in the post-Cold

War period than an attempt at social science” (Johnson 1997, 171). Indeed, it is somewhat ironic

for its detractors that such a culturally bound theory as rational choice is “sophomoric […] when



Fisher_DK3638_C013.indd 177



9/19/2006 3:47:41 PM



178



Handbook of Public Policy Analysis



contending that [it] contains a unique capacity to transcend culture and reduce all human behaviour

to a few individual motivational uniformities” (Johnson 1997, 1).

Typically, the alleged primacy of self-interest is said to fail to recognise the complex motivations of individuals in the policy process. As Udehn (1996) argues in an attack on the assumption

of self-interest, policy actors are just as likely to be motivated by group interests or altruistic public

concerns as their own narrow self-interest; and, as such, self-interest cannot be simply imputed on

or read into every action of an individual throughout the process of decision-making. Exploring

rationalist explanations of bureaucratic behavior, Udehn thus questions Niskanen’s characterization of bureaucrats (1971) as self-interested budget-maximizers, eager to increase departmental

budgets in order to provide themselves with opportunities to award themselves higher salaries,

gain promotion, or enhance their own reputation. For Udehn there is “no necessary connection

between budget-maximisation and self-interest […] the way to make a career might be to reduce

budgets or, at least, to save money” (1996, 75). Budget-maximisation might not necessarily be a

case of self-interested behavior, but an example of the altruistic desire to serve the common good.

In fact, the imposition of a utility function in rational choice accounts attributes similar motivations

to individuals such that context comes to dominate in explanations over and beyond the agency of

individual actors (as we have suggested above). As such, rational choice is charged with ignoring

the origins of preferences and over-riding the past history and identity of policy actors, stripping

individuals of differences that emerge from their individual personal identity or history; with no

account of identity-formation, actors in the policy process become interchangeable (Tsebelis 1990,

43). Whilst prized by mainstream rational choice theorists as a short-cut to scientific explanation,

this interchangeability is derided by detractors as a “homogeneity assumption” that “is justified on

the grounds of theoretical parsimony” (Green and Shapiro 1994, 17). It elevates the policy context or

institutional environment to the exclusion of individuals and choice, such that “it is surely tempting

to ask: ‘When is a choice not a choice?’ Answer: ‘When it is a rational choice.’” (Hay 2004, 53).

Such short cuts toward parsimonious and universal explanation are alleged ultimately to underplay the uncertainty and complexity of much of the decision-making process. Here, policy making

is far removed from the strategic cost-benefit calculations of informed and instrumental individuals.

Rather, it is characterized by the politics of ambiguity and uncertainty where the information available to policy-makers is inconclusive, where their capacity to process such information remains

limited, and where consequently policy making becomes less about problem-resolution and more

about persuasion, communication, advocacy, and the setting of norms (Majone 1989). Indeed,

uncertainty, unlike complexity and risk, hampers the definition by individuals of both interests and

distinct courses of action. It obliges actors to operate within policy frames that establish norms and

codes for making sense of problems and guiding their behaviour through the policy process (Griggs

and Howarth 2002). Under such conditions of ambiguity, individual policy actors are thus best conceived of as satisfiers rather than maximizers, operating under the conditions of bounded rationality

(Simon 1957) or following institutional logics of appropriateness (March and Olsen 1989).

In fact, methodological pathologies and inadequate empirical testing are, in the work of Green

and Shapiro (1994), said to fuel these untenable rationalist characterisations of the dynamics of

policy-making. Labelling rational choice as method rather than problem-driven, Green and Shapiro

do indeed argue that rational choice theory betrays a bias towards post hoc theory development

(1994, 34–39), epitomized by the process of retroduction whereby rational choice theorists deliver

post hoc explanations of known facts. Such post hoc theorizing undermines the empirical testing

of rational choice inspired models, for as they argue:

Data that inspire a theory cannot, however, properly be used to test it, particularly when

many post hoc accounts furnish the same prediction. Unless a given retroductive account

is used to generate hypotheses that survive when tested against other phenomena, little of

empirical significance has been established. (Green and Shapiro 1994, 35–36)



Fisher_DK3638_C013.indd 178



9/19/2006 3:47:41 PM



Rational Choice in Public Policy



179



Indeed, Green and Shapiro claim that this absence of empirical testing is amplified by the recourse

to a comparatively large number of unobservable entities in rational choice explanations (1994,

39–41). Thus, they posit that when rational choice models confront unexpected outcomes or patterns of behavior, they are able to dismiss such outcomes as some unobservable thought process

or “offsetting tendency or temporary aberration”; evidence for which it is difficult to establish

empirically (1994, 40).

What holds such methodological and theoretical challenges together is the argument that rational choice fails to address the empirical, both in its methodological pathologies, which constrain

its empirical testing, and in its parsimonious and universal assumptions over the motivations and

roles of individuals, which fail to conceptualise the complexity of the decision-making process.

Nowhere is this more clearly illustrated than in the debates surrounding the veracity of rational

choice theories of collective action. Here Olson’s problematization of collective action and the

requirement to offset the tendency of rational actors to free-ride through the generation of selective

pecuniary incentives are simply said to be refuted empirically by the range of established patterns

of group mobilization and social movement protest and contestation across political systems (see,

for example, the work of Jordan and Maloney 1997). Such forms of mobilization are not temporary

aberrations, but point arguably to the limitations of instrumental accounts in modelling the complex

motivations of individuals and the need for rational choice theory to address rival explanations of

group identity and its expression if it is to understand the limits to its universalist aspirations. It is

to the response of rational choice theorists to such challenges that we now turn.



MEETING THE CHALLENGES TO RATIONAL CHOICE

All research communities are composed of “pragmatists,” “synthesizers,” and ”competitors” (Lichbach 2003, 4–7). Whilst “pragmatists” work solely within the confines of their chosen approach,

“synthesizers” complement their chosen approach with the insights of rival theories in the desire

to converge towards a unified centre, and “competitors” advance the engagement and evaluation of

their chosen approach against other theories, stressing the incompatibility of fundamental competing assumptions between approaches. In this section, drawing on the work of Lichbach (2003),

we present three stylized accounts of such potential responses to the challenges to rational choice

theory. Protectionist pragmatists, the default position of much of rational choice theory, seek to ward

off challenge and maintain the fundamental building blocks of rational choice intact. Expansionist

synthesizers start from the assumption that rational choice should deliver an effective portrayal of

the reality of decision making and thus seek to rearticulate the assumptions of rational choice theory

to ward off claims of simplification and lack of complexity. Finally, confrontational competitors

privilege the fundamental assumptions of rational choice, but call for further engagement of rational

choice theory with its rival approaches based upon a dynamic testing of rival predictions; a process

that has the potential to reach a substantive amelioration of rational choice theory. Let us begin our

analysis by examining the responses of pragmatic protectionists.



PRAGMATIC PROTECTIONISTS

The standard defense of such pragmatic protectionists is to adopt what is termed the Friedman defence

(Friedman 1953). Here advocates of rational choice acknowledge that the simplifying assumption

of rationality in no way mirrors or represents the complexity of policy making. However, they argue

that this is somewhat immaterial for the strength of any theory lies not in the accuracy of its basic

assumptions, but rather in its predictive and explanatory power (Green and Shapiro 1994, 30–31;

Hay 2004, 48). And against these evaluative criteria, rational choice has demonstrated its value,



Fisher_DK3638_C013.indd 179



9/19/2006 3:47:42 PM



180



Handbook of Public Policy Analysis



not least through the problematization of collective action. In addition, however, such pragmatic

protectionist responses tend to draw upon the “covering law” defense. Here, rationalists deny the

complexity of the policy process and argue in fact that the assumption of rationality does indeed

represent the actual practice of decision-making (Green and Shapiro 1994, 30–31; Hay 2004, 48).

Finally, a further pragmatic protectionist defence is to invoke the defence of domain restriction.

Tsebelis typically argues that rationality is best defined as a “subset of human behaviour” (1990,

32). Thus rational choice cannot claim to explain all political phenomena. Rather, in a relaxation

of its universal aspirations, rational choice should be confined to episodes where “the actors’ identity and goals are established and the rules of interaction are precise and known to the interacting

agents” (1990, 32).

Such responses tend to keep the fundamental assumption of rationality intact in that they either

dismiss challenges as to the requirement of any theory of the policy process to portray the reality of

decision making, deny the complexity of the decision-making process, or limit the applicability of

rational choice to particular domains of decision-making. However, these pragmatic protectionist

rebuttals are far from complementary and draw upon contradictory justifications (see below). More

importantly, they all run the risk of ignoring disconfirming evidence, being biased in other words

towards the privileging of evidence that simply reinforces the assumption of rationality (Green and

Shapiro 1994, 42–43). Indeed, domain restriction abandons whole areas of the policy process to

alternative explanations, which is not without implications for both the applicability and empirical

testing of rational choice models: “If the appropriate domain within which a theory is to be tested

is defined by reference to whether the theory works in that domain, testing becomes pointless”

(Green and Shapiro 1994, 45).



EXPANSIONIST SYNTHESIZERS

Expansionist synthesizers, in what could be construed as a hegemonic move to expand the boundaries of rational choice (Lichbach 2003), rearticulate the fundamental assumptions of rationalist

accounts in order to offer an effective portrayal of the reality of decision making. In so doing, they

adopt to varying degrees the position of the expansionist synthesizer. Thus, on the one hand, Dunleavy (1991) enriches the work of Niskanen by refining the utility function of bureaucrats and the

conditions associated with the following of budget-maximizing strategies by bureaucrats, arguing

that budget-maximization will depend upon the position or rank of the bureaucrat within the bureau

and the nature of the budget that is to be maximised. Where conditions are not opportune, senior

bureaucrats will bureau-shape rather than budget-maximise. On the other hand, other rational choice

models begin to re-articulate conceptions of self-interest further away from the pursuit of narrow

pecuniary benefits in order to portray more effectively the reality of the motivations of individuals in

the decision-making process. In the words of Ward, such accounts remove the straitjacket of rational

choice and allow for the likes of bounded rationality, choice under uncertainty and nonegotistical

and “moral motivations” (Ward 2002, 84). Typically, work on collective action has thus come to

privilege the role of ‘soft’ selective incentives, such as expressive or social incentives that can only

been gained from participation (see the work of Chong (1991) and Opp (1986) and Opp and Gern

(1993)). Altruism thus becomes a dominant utility-function, although rational choice theorists have

recognized its importance in the past (Ward 2002, 79).

Such moves toward identity-based interpretative explanations are firmly endorsed by the work

of Ferejohn (1991). His starting point is to claim that both rational and interpretative accounts of

public policy are incomplete, unable on their own to offer a full account of social action. Indeed, a

complete account of social action, he argues, is only possible if we are to view rational and interpretative approaches as complementary (1991, 279). Rational choice is thus defined for Ferejohn



Fisher_DK3638_C013.indd 180



9/19/2006 3:47:42 PM



Rational Choice in Public Policy



181



as “an interpretive theory that constructs explanations by ‘reconstructing’ patterns of meanings and

understandings (preferences and beliefs) in such a way that agents’ actions can be seen as maximal,

given their beliefs” (1991, 281). Indeed, the logics of rational choice and interpretative approaches

are similar in that they are both intentional accounts that

start with observed data (behaviour including documents and letters, practices and institutions) and reconstruct actors and their inner attributes (meanings, beliefs and values)

in such a way that the data are as fully explained or accounted for as possible. (Ferejohn

1991, 281)

And, as for interpretative accounts, they require a form of rationality embedded in them (1991,

281). This expansionist account is thus that of a fully-fledged “synthesizer”

Underpinning such responses is the distinction between “thin” and ”thick” accounts of rationality. Mainstream rational choice accounts of policy change “posit not only rationality, but some

additional description of agent preferences and beliefs” (Ferejohn 1991, 282), imputing onto agents

a dominant utility function so that for example all politicians across different contexts will seek to

maximize votes whilst bureaucrats will seek to maximise budgets. However, “thin” accounts assume no more than instrumental accounts of rationality that agents “efficiently employ the means

available to pursue their ends” and impute no dominant utility function onto individuals (Ferejohn

1991, 282). Indeed, as we said above, what these accounts are doing is not so much loosening the

assumptions of rational choice, as questioning how far self-interest has to be defined in terms of

narrow “selfish” or hard pecuniary benefits. They thus focus upon the re-articulation of self-interest

as utility maximisation in order to problematize the rationalist conception of individuals as selfish

individuals. The conception of self-interest comes to mean simply that individuals pursue preferences

in pursuit of benefits that may or may not be “selfish” or material in nature. Indeed, utility replaces

the notion of self-interest, such that “formally speaking, self-interest is not required for rational

choice, utility maximisation can suffice” (Dowding and King 1995, 14). In short, what this means

in practice is that utility is to be privileged over self-interest, but also that the very conception of

utility itself is to be stretched. Such accounts run the risk of tautological explanation for there is no

course of action that cannot be explained away as some form of utility-maximisation once rational

choice accounts move away from the preoccupation with hard pecuniary benefits. Equally, claims

to universalism and prediction are further weakened.



CONFRONTATIONAL COMPETITORS

Negating the potential hegemonic drives of expansionists, confrontational competitors call for the

further engagement with, and testing of rational choice theory against, its rivals. Thus, Lichbach

(2003, 209) asserts that the “success of rational choice depends on how it treats its foils.” Rational

choice theory should tackle concrete problems, establishing sets of competing predictions against

which it can engage in “creative competitions” (2003, 209). This is not to negate the fundamental

assumptions of rational choice, for as Lichbach suggests (2003, 62):

Rational choice theorists should begin with thin theories (that assume pecuniary selfinterest) as their baseline model, determine how much of the phenomena under question

can be explained by such theories, and only then adopt thicker theories to account for

phenomena that remain inexplicable. […] But before they attempt to synthesize their

major competitors, rational choice theorists should initially remain consistent with the

core of rational choice theorizing.



Fisher_DK3638_C013.indd 181



9/19/2006 3:47:42 PM



182



Handbook of Public Policy Analysis



Rather it is ultimately to call for a more “modest modelling tradition” of rationalist narratives: “the

historical and comparative analysis of strategic decision-making’” (Lichbach 2003, 211).

In fact, through the construction of analytic narratives, Bates and colleagues (1998) bolster

rational choice explanations of events and outcomes through drawing systematic explanations from

“thick descriptive” case studies, which underline the importance of social context and history. In

doing so, Bates and colleagues move toward a problem-driven approach that employs deductive

reasoning and thin accounts of rationality, but retreats from the production of universal laws of

human behaviour (Bates et al. 1998, 11). Such an approach offers a narrative approach that both

examines “stories, accounts and context” and “extracts explicit and formal lines of reasoning,

which facilitate both exposition and explanation” (Bates et al. 1998, 10). It establishes the origins

of individual preferences and dominant rules and patterns of strategic interaction, building up from

the contextual reconstruction of events as observed and interpreted by actors. As such, it combines

inductive qualitative and descriptive methods to establish actors’ preferences and perceptions with

deductive methods to understand the behavior of actors within the game being analysed (Bates et

al. 2000, 697). The pursuit of analytic narratives thereby addresses the tendency of mainstream

rational choice to “take people as we find them” (see above), deploying narratives in the context of

actor-centric extensive-form games.

The employment of analytic narratives responds directly to challenges over the tendency of

rational choice accounts to impute beliefs and preferences onto actors in the policy process. Indeed,

the recourse to “thick descriptive” case studies becomes a central step in rational explanation, for

it produces in the words of Bates and his colleagues “such intimacy with detail” that “we argue,

must inform the selection and specification of the model to be tested and should give us a good

grasp of the intentions and beliefs of the actors” (Bates et al. 2000, 698). Like the approaches discussed above, however, the construction of analytic narratives is not without its criticisms. First, it

leads to the questioning of the actual value that the assumption of rationality adds to explanations.

Second, it is claimed that the approach is prone towards tautological explanations (Elster 2000).

For as Bates and his colleagues argue, “Should we possess a valid representation of the story, then

the equilibrium of the model should imply the outcome we describe—and seek to explain” (Bates

et al. 1998, 12).

Overall, therefore, rational choice has tended to generate three types of responses to the

challenges posed by alternative accounts or indeed generated by its own internal contradictions.

Whilst default responses have sought to keep the assumption of rationality broadly intact, other

responses have sought either to reinterpret the primary assumptions of rational choice to adopt a

more synthetic approach that brings together rational and interpretative approaches. What these

more expansionist accounts, as with confrontational accounts, deliver is the relaxation of the universalist and parsimonious aspirations of rational choice through the investigation of the origins of

preferences and individuals’ conceptions of utility. However, as we suggested above, we have to be

wary of the potential to confuse the different epistemological positions of instrumental-empiricism

and scientific-realism that underpin such accounts. Although often grouped together, the Freidman

inspired “as if” defense builds upon instrumental-empiricism whilst domain restriction is based

upon scientific-realism. These different epistemological positions in rational choice theory require

clarification (MacDonald 2003).

Equally, we have to guard against the metaphorical reduction of rational choice theory to the

status of a neutral or pliable toolkit within the armoury of policy analysts; albeit a more than useful

tool in the armoury open to analysts. Indeed, in pursing the development of analytic narratives, Bates

and his colleagues challenge Elster, one of their critics, to “offer us a better set of tools” (Bates et al.

2000, 702). This position is not without its detractors. Udehn argues that such a tactic of reducing

rational choice to a heuristic device is undoubtedly limited “unless it is argued that the failures of a

simplifying assumptions are as enlightening as its success” (1996, 93). However, rather than enter



Fisher_DK3638_C013.indd 182



9/19/2006 3:47:42 PM



Rational Choice in Public Policy



183



into further discussion of the validity of the assumption of rationality, we argue that the conception

of rational choice as a heuristic device or a “set of tools” misunderstands the constitutive nature of

policy theory. It is to this misunderstanding that we turn in our conclusion.



ASSESSING THE LIMITS OF RATIONAL CHOICE: IT IS WHAT IT IS, ISN’T IT?

Theories of the policy process constitute the social reality they seek to describe and to interpret,

offering particular explanations of the research problems and data they themselves have constituted

(Howarth 2000, 130). Policy theories do not exist outside of the data. This starting point places

limitations on rational choice theory and its responses to its internal and external challenges. First,

it questions the empiricist methodology of mainstream rational choice accounts and undermines

its claims to predictive universal explanation. Second, it challenges specifically recent attempts to

relegate rational choice to a toolkit or partial explanation to be deployed in conjunction with other

approaches, for these accounts of the limits of rational choice are predicated upon the existence of

empirical data that somehow exists as valid outside of theory and neglect how particular theories

selectively determine the definition of what is data and indeed what are the relevant research questions to be explored. In fact, this flaw can arguably be associated with the arguments of Green and

Shapiro in their seminal critique of rational choice theory. Green and Shapiro condemn rational

choice because in rational choice accounts “evidence is projected from theory rather than gathered

independently from it” (1994, 42). Thus, they themselves fail to acknowledge the constitutive logic

of policy theory. There is not a set of empirical data that can be generated by neutral tools independent of theory and that can then be subjected to different theoretical approaches. To even approach

explanation requires some initial or partial understanding (Howarth 2000, 131).

What does this mean for the limits of rational choice? In short, it offers an understanding of

rational choice that is best summarized as “it is what it is.” This is not to call for the absence of

dialogue between different approaches to the explanation of policy outcomes. Neither is it to necessarily adopt the Freidman defense or that of the covering law or domain restriction. Rather, it is to

argue that rational choice theory, like all policy theory, offers a particular explanation of research

problems and data it itself has constructed. It cannot be reduced to a neutral toolkit or heuristic device or integrated into other approaches. Its limitations are thus to be judged through debate within

the community of policy analysts through the likes of creative confrontations. After all, there is no

neutral body of data that can be constructed through the use of impartial tools and against which

we can test theories. Rational choice is what it is, isn’t it?



REFERENCES

Alt, J.E., and Chrystal, K.A. (1983). Political Economics. Brighton: Wheatsheaf Books.

Arrow, K.J. (1951). Social Choice and Individual Values. New York: Wiley.

Axelrod, R. (1984). The Evolution of Co-operation. New York: Basic Books.

Bates, R.H., Greif, A., Levi, M., Rosenthal, J-L., and Weingast, B.R. (1998). Analytic Narratives. Princeton,

NJ: Princeton University Press.

Bates, R.H., Greif, A., Levi, M., Rosenthal, J-L., and Weingast, B.R. (2000). The Analytic Narrative Project.

American Journal of Political Science, 94, 696–702.

Blyth, M. (2002). Great Transformations. Economic Ideas and Institutional Change in the Twentieth Century.

Cambridge: Cambridge University Press.

Chong, D. (1991). Collective Action and the Civil Rights Movement. Chicago: University of Chicago.

Chong, D. (1996). Rational choice theory’s mysterious rivals. In J. Freidman (ed), The Rational Choice Controversy, pp. 37–57. New Haven: Yale University Press.



Fisher_DK3638_C013.indd 183



9/19/2006 3:47:42 PM



184



Handbook of Public Policy Analysis



Dowding, K., and King, D. (1995). Introduction. In K. Dowding and D. King (eds), Preferences, Institutions

and Rational Choice, pp. 1–19. Oxford: Clarendon Press.

Downs, A. (1957). An Economic Theory of Democracy. Harper and Row: New York.

Downs, A. (1967). Inside Bureaucracy. Boston: Little Brown.

Dunleavy, P. (1991). Democracy, Bureaucracy and Public Choice. Economic Explanations in Political Science.

Hemel Hempstead: Harvester Wheatsheaf.

Elster, J. (1986). Introduction. In J. Elster (ed), Rational Choice, pp. 1–33. Oxford: Basil Blackwell.

Elster, J. (2000). Rational Choice History: A Case of Excessive Ambition. American Political Science Review,

94, 685–695.

Ferejohn, J. (1991). Rationality and Interpretation: Parliamentary Elections in Early Stuart England. In K.R.

Monroe (ed), The Economic Approach to Politics. A Critical Reassessment of the Theory of Rational

Action, pp. 279-305. London: Harper Collins.

Friedman, J. (1996). (ed). The Rational Choice Controversy. Economic Models of Politics Reconsidered. New

Haven: Yale University Press.

Friedman, M. (1953). The Methodology of Positive Economics. In M. Friedman (ed), Essays in Positive

Economics, pp. 3–46. Chicago: University of Chicago Press.

Garrett, G., and Weingast, B.R. (1993). Ideas, Interests, and Institutions: Constructing the European Community’s Internal Market. In J. Goldstein and R.O. Keohane (eds), Ideas and Foreign Policy: Beliefs,

Institutions and Political Change, pp. 173–207. Ithaca: Cornell University Press.

Green, D.P., and Shapiro, I. (1994). Pathologies of Rational Choice Theory. A Critique of Applications in

Political Science. New Haven: Yale University Press.

Griggs, S., and Howarth, D. (2002). The Work of Ideas and Interests in Public Policy. In A. Finlayson and J.

Valentine (eds), Politics and Post-Structuralism: An Introduction, pp. 97–111. Edinburgh: Edinburgh

University Press.

Goldstein, J., and Keohane, R.O. (eds) (1993). Ideas and Foreign Policy: Beliefs, Institutions and Political

Change. Ithaca: Cornell University Press.

Hardin, G. (1969). The Tragedy of the Commons. Science, 162, 958–962.

Hay, C. (2004). Theory, Stylized Heuristic or Self-Fulfilling Prophecy? The Status of Rational Choice Theory

in Public Administration. Public Administration, 82, 39–62.

Howarth, D. (2000). Discourse. Buckingham: Open University Press.

Johnson, C. (1997). Preconception vs. Observation, or the Contribution of Rational Choice Theory and Area

Studies to Contemporary Political Science. PS: Political Science & Politics, 30, 170–174.

Jordan, G., and Maloney, W. (1997). The Protest Business? Mobilizing Campaign Groups. Manchester: Manchester University Press.

Laver, M. (1997). Private Desires, Political Action. An Invitation to the Politics of Rational Choice. Sage:

London.

Levi, M. (1997). A Model, A Method, and a Map: Rational Choice in Comparative and Historical Analysis. In

M.I. Lichbach and A.S. Zuckermann (eds), Comparative Politics. Rationality, Culture and Structure,

pp. 19–41. Cambridge: Cambridge University Press.

Lichbach, M.I. (2003). Is Rational Choice Theory All of Social Science?, Ann Arbor: University of Michigan.

MacDonald, P.K. (2003). Useful Fiction or Miracle Maker: The Competing Epistemological Foundations of

Rational Choice Theory. American Journal of Political Science, 97, 551–565.

Majone, G. (1989). Evidence, Argument and Persuasion in the Policy Process. New Haven: Yale University

Press.

March, J., and Olsen, J. (1989). Rediscovering Institutions. New York: Free Press.

Mueller, D.C. (2003). Public Choice III. Cambridge: Cambridge University Press.

Niskanen, W. (1971). Bureaucracy and Representative Government. Aldine-Atherton: Chicago.

Nordhaus, W.D. (1975). The Political Business Cycle. Review of Economic Studies, 42, 169–190.

North, D.C. (1990). Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge

University Press.

Olson, M. (1965). The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA:

Havard University Press.

Olson, M. (1982). The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities. New

Haven: Yale University Press.



Fisher_DK3638_C013.indd 184



9/19/2006 3:47:43 PM