Chapter 12. Rationality in Policy Decision Making

162



Handbook of Public Policy Analysis



only the current consensus—sometimes very long-lived—within a community of researchers. Thus

science has a communicative context, and scientific knowledge claims are contingent.

Science is an elite enterprise that systematically excludes or marginalizes incompetent practitioners. Scientific norms, practices of peer review, and credentialing institutions all work to filter

out poorly trained and less talented researchers, although these arrangements can also encourage

a stifling orthodoxy. Democracy, it is worth noting, also has elements of exclusivity—madmen,

children, and noncitizens do not vote (Guston 1993). Yet it remains true that only a tiny segment

of society produces, communicates, and advocates on behalf of the scientific knowledge used in

public decision making. Its moral autonomy has “helped create science in an Old Testament style,

that is, respect without comprehension” (Toumey 1996, 34).

There are other sources of relevant, elite-based knowledge. The legal expertise of attorneys,

the moral authority of religious leaders, the sharp pencils of accountants, and the contributions of

many types of professionals all might contribute to better public decision making.

Difficulties emerge when deciding whose knowledge counts the most. Citizens vary in their

education, life experience, raw intelligence, and particular knowledge. It makes sense to delegate

certain decisions to experts. We ask engineers to design our bridges, neurosurgeons to operate on

our brains, and economists to set monetary policy. Yet sometimes the value of elite expertise is less

clear. Thus toxicological knowledge may be no more relevant than local fishing lore when setting

fish consumption advisories in polluted waters. Experts also have limited domains of expertise, so

it is not clear whether a biologist or a priest has a stronger claim in helping politicians decide the

ethics of embryonic stem cell research. Some scientific knowledge is uncertain or irrelevant, some

relevant knowledge is local rather than general, and some decisions turn on values as much as on

the facts.



SUBSTANTIVE AND PROCEDURAL DIMENSIONS

Simon (1976) offers one solution to the dilemma of whose knowledge counts by distinguishing between substantive and procedural rationality. He measures substantive rationality relative to optimal

outcomes, and procedural rationality relative to optimal decision processes.

Scientifically produced knowledge contributes to substantive rationality. Such knowledge describes phenomena and explains causal factors. It provides the factual basis for better decisions. It

satisfies scientific criteria of validity and reliability, and it justifies authoritative knowledge claims.

Scientific knowledge is a key input that contributes to the “best” outcome.

Yet every decision also has a procedural component: Who makes the decision, and what are the

prescribed steps in making the decision? Procedural rationality helps improve decisions by specifying

their “who” and “how” aspects. Optimal processes are legitimate, reasoned, and transparent.

Simon’s distinction between substantive and procedural rationality maps nicely onto a distinction made decades earlier by Weber (1922) regarding sources of legitimacy. Public decisions are

legitimate if they are legal, authoritative, and appropriate for the context. There are two very different

sources of legitimacy. First, there is the legitimacy of authority, based in status, as enjoyed by divine

rulers, revered elders, and scientific experts. Second, there is the legitimacy of consent, based in civil

society, and deriving from following constitutional rules and open, democratic procedures. These

distinctions make intuitive sense, but they also raise a troubling question: Are these two sources of

legitimacy complementary or does one displace the other? Symmetrically, does increased procedural

rationality come at the expense of decreased substantive rationality?

Lindblom (1990) argues forcefully that substantive rationality should not be allowed to displace

procedural rationality. Especially regarding social phenomena, he doubts that science ever rises

above mere confirmation of common sense. He believes that social scientific expertise should not



Fisher_DK3638_C012.indd 162



10/16/2006 10:49:00 AM



Rationality in Policy Decision Making



163



be privileged over lay judgment in public decision making, and he advocates a self-guided rather

than a scientifically guided society. Science should play only a minor supporting role. Lindblom’s

argument draws strongly on his famous “muddling through” conception, which asserts that scientific planning frequently fails in decentralized capitalist democracies because power is shared and

no one has full information (1959). His solution is to encourage incrementalism in policy-making,

and continual, mutual adjustment among policy actors.

Critics of Lindblom assert that incrementalism is the rational procedure sometimes but not all

of the time (Breheny and Hooper 1985; Smith and May 1980). There are decisions in which expert

contributions to substantive rationality are essential, such as the bridge-building example mentioned

earlier. There are also “big” decisions that cannot be easily or cheaply reversed or subdivided, such as

damming a major river, starting a nuclear power plant, or going to war. In such cases, the goodness

of decisions depends equally on their procedural and substantive elements. However, the critics face

a daunting task in explaining which decisions should be incremental and which should not.



RATIONAL AND REASONABLE

There is a way to relieve some of the inconsistencies identified in the preceding paragraphs, by

distinguishing between decision making and analysis. Decision making is the art of choosing reasonable decision rules, ones that are appropriate for each decision context. Reasonable decision rules

are internally consistent and an outcome of moral argumentation, they are values-based. Analysis is

the science of applying those decision rules rationally, according to appropriate standards. Rational

analysis is logical, valid, reliable, and empirically tested, it is fact-based. Values and facts are distinct

intellectual categories although they can be hard to distinguish from one another in practice.

“Irrational” decisions are often nothing more than the rational application of unreasonable

decision rules. For example, given a “net social benefit” decision rule, it may seem quite rational to

dam a major river like the Yangtze. If a “no losers” test had been applied instead, then the rational

choice might be to leave the river alone. Rationality in policy decision making thus requires both

reasonable decision rules and rational applications of those rules.

Often the jobs of analysis and decision making are separated because of the scale and complexity of modern society. This opens up the possibility that analysts will not check with decision

makers to determine which decision rule is reasonable. Equally, decision makers may not bother

to check with analysts regarding the rationality of the rule’s particular application. Rationality in

policy decision making thus depends crucially on effective communication between decision makers and analysts.

The troubling word “appropriate” appears twice in this section, raising difficult questions. How

does a decision maker know that a decision rule is reasonable, that is, appropriate to its context?

Also, how does an analyst know that the logical and empirical standards applied are rational, that is,

appropriate to the decision rule and its context? The answer to both questions is that appropriateness

is socially determined. Both decision makers and analysts thus must communicate with others—the

population at large as well as a range of experts—to confirm that they have acted appropriately.



SINGLE AND MULTIPLE DECISION MAKERS

Some decisions seem private: you choose your breakfast items, your day’s clothes, and whether to

get out of bed in the morning. Yet even despots do not get to make truly independent decisions. If

you know that others strongly influence your own choice of food, clothing, and schedule, you also

suspect that even Adolf Hitler and Julius Caesar needed to consult with confidantes and mollify the



Fisher_DK3638_C012.indd 163



10/16/2006 10:49:00 AM



164



Handbook of Public Policy Analysis



masses. Almost all policy decisions are truly multi-party decisions. This implies, first, that there

must be a procedural component to public decision making that specifies how the parties interact.

Second, this multi-party context implies that communication among parties will strongly influence

outcomes.

As the next section illustrates, a close look at the tools of rational policy analysis shows that

some of the most widely used methods blithely assume a single omniscient decision maker. For

example, a hypothetical social planner acts on behalf of the whole population when selecting socially

optimal policies by means of benefit-cost analysis. The risk assessor likewise weighs expected aggregate social risks and benefits when selecting optimal regulations. Such tools simplify away the

procedural and communicative challenges of public decision making. We need to move beyond

them (Andrews 2002).



TOOLS OF RATIONAL POLICY ANALYSIS

The science of public decision making has in fact evolved rapidly over the past century. A quick

tour spanning welfare maximization, public choice, multi-agent simulation, and decision support

follows.



WELFARE MAXIMIZATION

Starting from a microeconomic notion of individual utility, rationally pursued, Bergson, later extended by Samuelson, developed an additive social welfare function, with optimal public policy being

that which maximizes the aggregate utility across members of the population. This individualistic

formulation contrasts with competing concepts such as Benthamite utility (“the greatest good for

the greatest number”).

Benefit-cost analysis is a widely used tool that attempts to guide policymakers to welfare-maximizing choices. At its heart is the Kaldor-Hicks decision criterion that directs the “social planner”

to choose the alternative providing the greatest net social benefits. It depends on two very strong

assumptions, first, that it is reasonable to add gains and losses across individuals when calculating

the societal net benefit; and, second, that the things individuals gain or lose are easily substitutable

so that the winners can, at least hypothetically, compensate the losers. Thus it offers substantive

equivalence to the Pareto criterion that underlies the microeconomic conception of free markets,

because both can yield efficient outcomes.

However, the Pareto criterion also has a procedural component that promotes fairness as well

as efficiency: it requires unanimous and voluntary participation in transactions to ensure that decisions have no losers, only mutual gains. The Kaldor-Hicks criterion is more procedurally coercive,

implying majority rule at best, and dictatorship by the social planner at worst. Yet, realistically,

many things that civilizations need impose costs on a few in order to reap broad social benefits.

Examples include public education, progressive income taxation, highways, power plants, sewage

treatment plants, and jails.

Benefit-cost analysis thus has a place in the policy analysis toolkit. As a decision aid, it is

teachable, replicable, quantitative, and offers progressive insights that help narrow choices. Done

properly, it requires a comprehensive enumeration of costs and benefits—this is both a strength and

a weakness, since comprehensiveness is technically impossible even if it is conceptually desirable.

Practitioners have attempted to include an ever-broader range of tangible, intangible, and temporally

distant costs and benefits.



Fisher_DK3638_C012.indd 164



10/16/2006 10:49:00 AM



Rationality in Policy Decision Making



165



PUBLIC CHOICE

The microeconomic approach has led to a variety of contributions under the rubric of public choice

theory that show how individual interests influence both marketplace and public policy outcomes.

Self-interested bureaucratic actors maximize their budgets (Niskanen 1971), and citizens form clubs

to collectively provide some forms of public goods (Buchanan 1965), for example. The plausibility of social welfare functions has been challenged by Arrow (1951) on logical grounds, because

under reasonable assumptions it is possible to show that any democratic aggregation process leads

to inconsistent and unstable results.

Our collective decision-making mechanisms are deeply flawed, so that neither markets nor

politics on their own serve us adequately. Markets are sometimes inequitable because they concentrate wealth while making one dollar equal one economic vote, and because future generations get

no votes. Also, market imperfections abound, with monopoly power, public goods, externalities,

and information problems being the most severe weaknesses. Government failures also abound.

Bureaucratic self-interest can encourage agents within government to maximize their budgets or

allow regulatory agencies to be captured. Also, as discussed earlier, democratic decision making

according to “reasonable” criteria can become arbitrary and unstable (Arrow 1951). Given criteria

of collective rationality, a Pareto (no losers) principle, independence of irrelevant alternatives,

and non-dictatorship, Arrow presents an impossibility theorem for the existence of social welfare

functions. He shows that pairwise voting such as is currently practiced in most elections leads to

endless cycling, majority rule can select no winner non-arbitrarily, society’s preferences do not

exhibit collective rationality and do not aggregate individual preferences consistently, and that by

not counting the intensity of preferences, voting will lead to inconsistent outcomes (Pearce 1986).

Formal democratic decision-making mechanisms thus need to be supplemented sometimes.

One solution is to relax some of Arrow’s reasonable criteria. By relaxing the Pareto postulate, we open up the solution domain to allow winners, losers, and compensation. By relaxing the

independence postulate, we move from pairwise comparisons to multiple options. By relaxing the

unlimited domain postulate (part of collective rationality), we allow efforts to foster unanimity

prior to a vote. We maintain transitivity (the other part of collective rationality) and nondictatorship.

Public participation is one way to relax Arrow’s criteria while preserving elements of democracy,

as will be discussed shortly.

There have been many important extensions to the microeconomic approach that have broadened its reach. Game theorists have usefully characterized responses to governmental interventions

as mixed-motive (competitive-cooperative) negotiations (Axelrod 1984). Decision theorists have

introduced concepts of risk preference and multi-criteria tradeoffs (Keeney and Raiffa 1976),

thereby forcing into daylight the issues of risk and priority associated with policies. Psychologists

have demonstrated that individuals rely on heuristics and suffer from systematic biases that are

not reflected in standard utility functions (Tversky and Kahneman 1974), thereby weakening the

policy optimality claims of microeconomic analysts, and highlighting the roles of communication

and perception in public management. Most of these contributions have the effect of making the

science more effective in describing, explaining, and prescribing changes in public decision making,

in accord with an individualistic, utilitarian form of rationality.

Modern microeconomic theory now explicitly acknowledges both the independence and interdependence of decision makers. Theory now has much to say about the importance of rules and

access to information, pinpointing conditions leading to stable, efficient, and (sometimes) equitable

outcomes. Equally important, an emerging emphasis on collection of experimental evidence has

made microeconomics a more realistic behavioral science.



Fisher_DK3638_C012.indd 165



10/16/2006 10:49:00 AM



166



Handbook of Public Policy Analysis



MULTI-AGENT SIMULATION

Multi-agent simulation is a relatively new modeling approach that seeks to generalize from microeconomic game theory by incorporating more actors, imbuing them with more realistic cognitive

limitations in the form of bounded rationality and imperfect knowledge, and investigating out-ofequilibrium conditions. Access to cheap computing power has allowed modelers to explore cases

that used to be mathematically intractable, involving heterogeneous actors, evolutionary processes,

and emergent structural relationships. In policy simulations, preferred governmental interventions

sometimes diverge dramatically from those identified as optimal under neoclassical microeconomic

assumptions.

The multi-agent approach has been useful in the innovation, anti-trust, environmental, and

security policy domains, among others (Axelrod 1997; Barnett et al. 2000; Epstein and Axtell 1996;

Gilbert and Troitzch 1999). One early application shows how a widespread individual preference

for having as few as one-third of your neighbors share your own ethnicity will lead to highly segregated housing patterns, even in the absence of discriminatory real estate market actors and policies

(Schelling 1978). Another application shows how modeling with adaptive agents can yield results

that are at variance with those of neoclassical economics. In particular, in perfectly competitive

markets, regulation is either distorting and inefficient or irrelevant. However, when agents—firms

in this model—are heterogeneous, boundedly rational, and interact directly with one another out of

equilibrium, regulations can be demonstrated to have positive welfare effects (Teitelbaum 1998).



DECISION SUPPORT SYSTEMS

Another important innovation has been to reframe policy analysis as a decision support activity

rather than a surrogate for actual decision making. The distinction is important. Traditional policy

analysis accepts a request for a study from the decision makers, sets the scope of the research question, establishes key assumptions, carries out the analysis, and delivers policy recommendations

back to the decision makers. Decision support systems instead reserve key decisions for decision

makers, and plan for repeated interactions between analysts and decision makers. It is a humbler

approach but its analytics are often more complex.

An example of a decision support system is a regional electric power planning tool used a

few years ago in New England (Andrews 1992). Its purpose was to help break an impasse among

utility companies, regulators, environmentalists, and other interested parties regarding public utility investment policy. Analysts built a complex scenario analysis tool for simulating the operation

of the regional power system under various assumptions over a multi-decade time horizon, and

then they convened a planning process that involved all of these parties. Iteratively, the analysts

proposed the scope of analysis and the stakeholders approved it, the analysts offered assumptions

for review and the stakeholders approved them, when stakeholders disagreed among themselves

about certain assumptions the analysts ran their model both ways and reported back whether their

answers diverged, the stakeholders suggested policy alternatives and the analysts evaluated their

multi-attribute impacts, the stakeholders evaluated these simulated impacts and drew their own

conclusions about which solution was optimal from their point of view, and out of this interaction

emerged broad consensus on a new set of energy policies for the region. Decision support systems

have found application in an increasing number of fields spanning urban planning, environmental

policy, health policy, energy policy, international relations, and military policy, among others (Sauter

1997; Brail and Klosterman 2001).

Readers will detect that the key distinction between traditional policy analysis and decision

support systems is that the latter have an explicit procedural component. Decision support analysts



Fisher_DK3638_C012.indd 166



10/16/2006 10:49:00 AM



Rationality in Policy Decision Making



167



devote much effort to mapping the points in a decision process at which analysts might helpfully

intervene. They then apply themselves to the design of potentially fruitful interactions at those few

key points. They are shifting the practice of policy analysis away from a major focus on substantive

rationality to giving equal consideration to procedural rationality.

In sum, the tools of rational policy analysis have evolved over a century toward a richer conception of rationality that acknowledges substantive and procedural dimensions, rational applications

of reasonable decision rules, an expectation that interested parties hold diverse views, and more

limited roles for experts. These tools must appropriately blend facts and values as they produce

actionable policy advice.



PUBLIC PARTICIPATION

Public participation has several potential roles in policy decision making, and the reasons offered

for encouraging it are diverse (Wengert 1976). One can pursue participation as policy, that is, with a

normative perception that it is desirable in and of itself because, in the words of Susskind and Elliott

(1983), it democratizes, decentralizes, deprofessionalizes, and demystifies public policy. One can

pursue participation as a strategy, as a means for achieving other ends. Participation can operate as

communication, leading to improved information flows that produce better decisions. Participation

can serve as therapy, a way to coopt alienated groups into the mainstream. It can function as conflict

resolution, so that participation may (or may not) lead to reduced tensions and stable outcomes in

controversial decisions. More intuitively, the best way to find out what people want and value is to

ask them (Feiveson et al. 1976).

There is a spectrum of participation mechanisms available (FEARO 1988):

•

•

•

•

•

•



Public information (ads, newsletters, exhibits)

Public information feedback (polls, focus groups, surveys)

Consultation (hearings, workshops, panels, games)

Extended involvement (advisory committees, charrettes, task forces)

Joint planning (arbitration, conciliation, mediation, negotiation, partnership)

Delegation (citizen control, home rule)



Choosing which mode of participation to solicit is important because each has distinctive

strengths and weaknesses. Steps in choosing a mechanism typically include informal early consultation to identify major issues and actors, confirmation of an organizational mandate to proceed in an

accountable and open fashion, identification of potentially interested participants including those

who are unlikely to receive representation, analysis of the specific situation, setting objectives for

the participation process, determining information exchange requirements, planning the length of

the activity and the complexity of involvement, implementation, and evaluation of how it worked

(FEARO, 1988).

Cultivating public participation as a conscious strategy or policy does not mean that it may

not arise “naturally” in spontaneous occurrences. The existing pattern of public interaction sets the

stage. Susskind and Elliott (1988) describe patterns of paternalism (elites govern, they inform the

public of decisions), conflict (members of the public begin to distrust the elites, they second-guess

decisions and demand participation), and coproduction (elites and members of the public share

decision-making power and constructively resolve conflicts).

Like other collective decision-making mechanisms, public participation also has flaws. These

include potential tyranny of the majority or minority, instability of decisions, poor information,

apathy and stakeholder fatigue, adverse reactions to perfunctory involvement, and the nonrepresentativeness of participants in any small group process.



Fisher_DK3638_C012.indd 167



10/16/2006 10:49:01 AM



168



Handbook of Public Policy Analysis



In evaluating the success of public participation processes, Webler and Renn (1995) argue

that the key criteria are fairness and competence. A fair process gives all interested parties equal

opportunities to participate in the discourse, make validity claims, challenge validity claims made

by others, and determine when closure is reached. A competent process operates in direct tension

with fairness, however, by setting minimal standards for cognitive and lingual competence, ensuring reliance on appropriate knowledge, using a consensually approved translation scheme, and

relying on the most reliable methodological techniques available. In addition to balancing fairness

and competence, traditional measures of effectiveness and value also figure into perceived success.

Participation increases the rationality of public decision making to the extent that it overcomes the

failings of markets and politics, as well as its own internal weaknesses.



THEORETICAL AND PRACTICAL REASON

Philosophers identify theoretical and practical reason as two distinct approaches to the facts-values

dichotomy in public decision making. Theoretical reason is the rationality underlying pure science,

and it helps us evaluate whether our theories have empirical validity. Practical reason is the rationality underlying decision making, and it helps us evaluate the normative validity of our actions.

Marrying facts and values—analysis and decision making—requires integrating theoretical and

practical reason.

Is such a marriage necessary or even desirable? Perhaps only one type of reason is enough.

Theoretical reason, for example, is a consequentialist, outcome-oriented doctrine that prizes usefulness in linking facts and values. For its disciples, good science generates testable hypotheses and

credible data that usefully advance knowledge, and good decision making leads to useful, utilityenhancing outcomes.

Practical reason, by contrast, is a deontological, obligation-oriented doctrine that prizes wide

acceptance in linking facts and values. For its disciples, good science produces widely accepted

knowledge, and good decision making is also widely accepted.

Unfortunately, neither theoretical nor practical reason offers truly universal lessons. Theoretical

reason delivers only temporary truths because science is always contingent. Practical reason delivers only useful norms because experience is never adequate to support the formulation of absolute

moral laws. Both doctrines offer plausible ways to integrate facts and values in decision making,

as long as one is willing to accept an extremely limited, relative form of rationality.



INSTRUMENTAL, STRATEGIC, AND COMMUNICATIVE RATIONALITY

Jantsch (1975) points out that different contexts demand different rationalities. When the task is

simply to rationalize the use of scarce resources in the absence of uncertainty or conflict, an instrumental response is appropriate. Instrumental rationality optimizes the allocation of resources

according to an efficiency criterion. For example, a public works manager operates instrumentally

to allocate a fixed budget among pipes, valves, filters, pumps, concrete, excavation equipment, and

labor when designing a water supply system.

When the task is to rationalize a set of steering principles for managing uncertainty and complexity, a strategic response is appropriate. Strategic rationality optimizes strategies for responding

to change induced at least partly by the actions of others, according to an effectiveness criterion.

For example, a mayor operates strategically when deciding whether to build a water supply system

ahead of or in response to demand, even as developers and prospective residents decide whether to

build and buy new homes in her city.



Fisher_DK3638_C012.indd 168



10/16/2006 10:49:01 AM



Rationality in Policy Decision Making



169



When the task is to rationalize a set of collective preferences—norms—for managing conflict,

then a communicative response is appropriate. Communicative rationality optimizes opportunities

for achieving consensus, according to a criterion of ethical behavior. For example, the mayor may

ask a municipal planning commission to develop, in consultation with stakeholders and the general

public, a widely accepted land use plan that represents the community’s vision of where and when

future development will happen, including any required water supply system improvements.

A devotee of theoretical reason, say, a utilitarian, will respond instrumentally in a non-social

context and strategically in a social, multi-party context. This person will test the validity of scientific

facts by considering the empirical evidence. She will test the validity of decision-making values by

confirming that her utility increases.

A devotee of practical reason, by contrast, will respond instrumentally in a non-social context

but she will act communicatively in a social, multi-party context. She will test the validity of both

facts and values with reference to the degree of consensus each enjoys.

So, in a non-social context, devotees of both theoretical and practical reason operate instrumentally. But in a social context—which is far more prevalent—there is divergence because a theoretical

reasoner will operate strategically and a practical reasoner will operate communicatively (see Table

12.1). Thus, if the mayor is a theoretical reasoner, she will extend her town’s water supply system

strategically, seizing opportunities as they arise, without consultation, and maximizing her own

expected benefits along the way. On the other hand, if she is a practical reasoner, the mayor will

operate in a more deliberative manner, seeking broad input and acceptance of the water system’s

expansion plans and ensuring that neither she nor developers and residents are blind-sided by any

aspect of the development process.

Such an argument overstates the differences, however, because there can be utilitarian reasons

to operate communicatively. First, conflict can be a source of scarcity and uncertainty, giving communication both instrumental and strategic value. Second, the contingent and socially constructed

nature of scientific knowledge implies communication challenges, and highlights the value of

achieving consensus on “serviceable facts.” Third, since every public decision has a procedural

component, interested parties will have opportunities to reward conformance with widely shared

norms of ethical behavior. Fourth, the contemporary common division of labor between analysts

and decision makers can impose costly communicative distortions. Distortions also can appear

as experts communicate across disciplinary boundaries, or with members of the lay public. Thus,

communicative action offers utility for avoiding unintended misunderstandings and needless conflicts. All public decisions need to achieve at least a limited degree of communicative rationality,

as every mayor knows.



TABLE 12.1

Rationality in Philosophical and Social Context

Social Context:

Philosophical Context:



Single Party

Decision Context



Multi-Party

Decision Context



Utilitarian

Perspective

(consequentialist,

optimizing,

theoretical reason)



Instrumental Rationality

(rationalize use of scarce resources,

valid facts are empirically tested,

valid values increase utility)



Strategic Rationality

(rationalize steering principles given

uncertainty, valid facts are empirically

tested, valid values increase expected utility)



Communicative

Perspective

(deontological,

consensus-seeking,

practical reason)



Instrumental Rationality

(rationalize use of scarce resources,

valid facts are empirically tested,

valid values increase utility)



Communicative Rationality

(rationalize norms for managing conflict,

valid facts enjoy consensus,

valid values enjoy consensus)



Fisher_DK3638_C012.indd 169



10/16/2006 10:49:01 AM



170



Handbook of Public Policy Analysis



Norms of ideal communication include comprehensible, sincere, legitimate, and truthful speech

(Forester 1980; Habermas 1979). A typology of communicative distortions distinguishes two dimensions: origins and intentions. Individual and unintentional distortions include poor writing or

speaking capabilities, for example. Individual and intentional distortions include such acts as lying

and deception. Systemic and unintentional distortions are due to such problems as miscommunication across disciplinary and organizational boundaries. Systemic and intentional distortions are

particularly problematic, and examples include propaganda and misleading advertising. Reducing

any of these distortions directly increases communicative rationality.

Conflicts nevertheless persist between devotees of theoretical and practical reason because the

underlying moral values differ. For the same reasons that we see the tragedy of the commons play out

in many policy domains, so we see that strategic action frequently displaces communicative action.

Individuals often have too much to gain by strategic action not to pursue it, even if it damages the

public good. Consensus has limited value in a world with great scope for independent action.



CONCLUSIONS

The desire to apply rationality to public decision making is a modern desire, bundled in with other

tenets of modernity. These include a faith that humankind is making qualitative progress, that there

is a definable “public good,” that individual actions matter, and that human inventiveness will create

more good things than bad things. These tenets are by no means universally accepted. Some see a

grimmer world in which public decision making is purely a pluralistic game, a play of power, an

unequal contest whose outcome is determined by social or economic structure, or a simple series of

do-able deals. Others see a world in which values are the only essential element of decision making,

and the factual element is secondary and perhaps even indeterminate. Rationality in public decision

making will be of little interest to partisans and demagogues, and efforts to improve the scientific

or procedural basis of public decisions are unlikely to win their support.

This chapter began by linking rationality in public decision making with science, with substance

and logic. It then added a procedural dimension that greatly enriched our notion of rationality, by

bringing in communicative and legitimacy concerns. It showed that popular tools of policy analysis

such as benefit-cost studies are deficient when judged against this broader definition of rationality; hence, better tools are under development. Rationality may even be improved by resorting to

procedural solutions such as public participation. Ultimately, rationality is a characteristic of public

decision making that some people desire some of the time. It has no guaranteed seat at the table. Its

very vulnerability should make us value it and pursue it all the more passionately.



REFERENCES

Andrews, C.J. (1992). Sorting out a consensus: Analysis in support of multi-party decisions. Environment and

Planning B: Planning and Design 19(2): 189–204.

Andrews, C.J. (2002). Humble Analysis. Westport, CT: Praeger.

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

Axelrod, R. (1984). The Evolution of Cooperation. New York: Basic Books.

Axelrod, R. (1997). The complexity of cooperation: Agent-based models of competition and collaboration.

Princeton, NJ: Princeton University Press.

Barnett, W.A., Chiarella, C., Keen, S., Marks, R. and Schnabl, H., eds. (2000). Commerce, Complexity, and

Evolution: Topics In Economics, Finance, Marketing and Management, Proceedings of the Twelfth

International Symposium in Economic Theory and Econometrics, International Symposia in Economic

Theory and Econometrics. Cambridge: Cambridge University Press.

Brail, R., and Klosterman, R., eds. (2001). Planning Support Systems. Redlands, CA: ESRI Press.



Fisher_DK3638_C012.indd 170



10/16/2006 10:49:01 AM



Rationality in Policy Decision Making



171



Breheny, M.J., and A.J. Hooper, eds. (1985). Rationality in Planning. London: Pion Limited.

Buchanan, J. (1965). An Economic Theory of Clubs. Econometrica 32: 1–14.

Epstein, J.M., and Axtell, R. (1996). Growing artificial societies: Social science from the bottom up. Cambridge, MA: MIT Press.

Federal Environmental Assessment Review Office (FEARO). (1988). Public Involvement: Planning and Implementing Public Involvement Programs. Prepared by Praxis, Inc. for FEARO, Hull, Quebec, Canada.

Feiveson, H.A., Sinden, F.W., and Socolow, R.H., eds. (1976). Boundaries of Analysis: An Inquiry into the

Tocks Island Dam Controversy. Cambridge, MA: Ballinger.

Forester, J. (1980). Critical Theory and Planning Practice. APA Journal 46(3): 275–286.

Gilbert, N., and Troitzsch, K.G. (1999). Simulation for the Social Scientist. Milton Keynes: Open University

Press.

Guston, D.H. (1993). The essential tension in science and democracy. Social Epistemology 7(1): 3–23.

Habermas, J. (1979). Communication and the Evolution of Society. Boston: Beacon Press.

Jantsch, E. (1975). Design for Evolution. New York: Braziller.

Jasanoff, S. (1990). The Fifth Branch: Science Advisors as Policymakers. Cambridge, MA: Harvard University

Press.

Keeney, R., and Raiffa, H. (1976). Decisions with Multiple Objectives: Preferences and Value Tradeoffs. New

York: John Wiley.

Lindblom, C.E. (1959). The science of ‘muddling through.’ Public Administration Review 19: 78–88.

Lindblom, C.E. (1990). Inquiry and Change. New Haven: Yale University Press.

Merton, R.K. (1973.) The Normative Structure of Science. Reprinted in Merton’s The Sociology of Science.

Chicago: University of Chicago Press.

Nash, J.F. (1950). Equilibrium points in N-Person Games. Proceedings of NAS.

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

Pearce, D. (1986). The MIT Dictionary of Modern Economics, entry on social welfare functions. Cambridge,

MA: MIT Press.

Sauter, V.L. (1997). Decision Support Systems. Hoboken, NJ: John Wiley and Sons.

Schelling, T.C. (1978). Micromotives and Macrobehavior. New York: W. W. Norton.

Simon, H.A. (1976). Administrative Behavior: A Study of Decision-making Processes in Administrative Organization. New York: Harper & Rowe.

Smith, G., and May, D. (1980). The Artificial Debate Between Rationalist and Incrementalist Models of

Decision-Making. Policy and Politics 8: 147–161.

Susskind, L., and Elliott, M. (1983). Paternalism, Conflict, and Coproduction. New York: Plenum Press.

Teitelbaum, D. (1998). An Adaptive Agent Approach to Environmental Regulation. Ph.D. disseratation, Pittsburgh, PA: Carnegie Mellon University.

Toumey, C.P. (1996) Conjuring Science. New Brunswick, NJ: Rutgers University Press.

Tversky, A., and Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science 185:1124–

1131.

von Neumann, J., and Morgenstern, O. (1944). Theory of Games and Economic Behavior. Princeton, NJ:

Princeton University Press.

Wallace, W. (1971). The Logic of Science in Sociology. New York: Aldine deGruyter.

Weber, M. (1922, 1957.) The Theory of Social and Economic Organization. New York: Free Press.

Webler, T., and Renn, O. (1995). A brief primer on participation: philosophy and practice. In O. Renn, T.

Webler and P. Wiedemann, eds. Fairness and Competence in Citizen Participation. Dordrecht: Kluwer

Academic Publishers, pp. 17–33 (Ch. 2).

Wengert, N. (1976). Citizen Participation: Practice in Search of a Theory. In A. Utton, ed., Natural Resources

for a Democratic Society. Boulder, CO: Westview Press, pp. 1–40.



Fisher_DK3638_C012.indd 171



10/16/2006 10:49:01 AM



Fisher_DK3638_C012.indd 172



10/16/2006 10:49:02 AM