Exceptional proff only 2

The value of the issue context approach for

scientific policy advice

Berry Tholen

Department of Public Administration, Radboud University Nijmegen, The Netherlands; Email: [email protected].

Abstract

Scientific policy advice, and research for such advice, involves making a broad set of choices, for

instance on research strategy and on the presentation of results. The issue context approach offers

guidelines for individual advisors and advisory bodies (and their designers) to deal with these deci-

sions. It distinguishes types of problems and connects particular strategies to each type. The issue

context approach is critically assessed in this paper. Can this approach be of help for responsible

scientific advisors and for designers of advisory bodies? We conclude that, for empirical and logical

reasons, the issue context approach cannot live up to its promise. This assessment helps to deter-

mine the limits of the ambitions of the public administration to make policy-making more scientific.

Key words: scientific advice; problem types; values in research.

1. Introduction

Scientific advisors have to make all kinds of choices. What kind of

investigation one has to become engaged in, whether one should

concentrate on theory testing or on finding new perspectives, what

variables should be included, how one should present the findings

and to whom etc.? When such issues are not left to individual scien-

tific advisors but are inscribed in the work processes of advisory

bodies, parallel issues confront the designers of scientific advisory

bodies. What procedures should be followed to develop research

questions, what type of advice should be provided, should it contain

researchers from different disciplines or not, should particular

groups of citizens be involved etc.? (On the questions for researchers

see Douglas (2009), Kitcher (2001, 2011), Lacey (2005) and

Longino (2002, 2004). On the issue of design principles for scientific

advisory bodies see Collins and Evans (2007) and Lentsch and

Weingart (2011)).

The answers that are given in the literature to these choices differ

hugely. As to the role of the scientific advisor, in line with Max

Weber (Weber 1958 (original 1918)) some claim that a scientific ad-

visor should merely provide inconvenient facts to policy-makers.

Others maintain, with Wildavsky (1979: 10, 13), that advisors

should come up with feasible policies and use all their rhetorical

skills to persuade policy-makers. On another issue some claim that

scientific advisors should always try to involve lay people in the pro-

cess of research and advice (Flyvbjerg 2001; Wagenaar 2011).

Others point out the complications and negative effects of such a

strategy (Brown 2009; Sarewitz 2011).

The diversity of contradicting positions and arguments suggests

that we should not look for general answers to these choices, but

should distinguish between particular settings or types of problems.

An elegant and promising solution of this kind is brought forward

by a particular approach in recent literature. According to this ap-

proach design principles for scientific advisory bodies and the roles

and choices of individual scientific advisors should mirror the exi-

gencies of the issue context of the policy problems at hand. More

particularly, the level of societal consensus on the issue and the de-

gree to which robust knowledge exists ideally determine the choices

of scientists and the type of involvement of citizens (Pielke 2007;

Hoppe 2010; Letsch and Weingart 2011).

This paper will concentrate on a critical assessment of the issue

context approach. First, it asks whether, or to what extent, the pre-

sented guidelines for scientific policy advisors and for a context-mir-

roring design of advisory bodies are sufficiently clear be used in

practice. Second, it asks whether, or to what extent, the arguments

for these guidelines are convincing.

The remainder of this paper is organized as follows: Section 2

provides an overview of the issues that scientific advisors (and

designers of advisory bodies) have to deal with. The issue context

approach promises to be of help with these issues, or at least some

of them. Section 3 spells out the central characteristics of this

approach. In Section 4 we turn to the actual assessment of the ap-

proach. Section 5 considers whether the issue context approach can

live up to its promises.

2. The issues a scientist policy advisor has to deal with

Lacey (2005) provides a convenient overview of the choices that sci-

entific advisors have to make. Researchers have to decide: what re-

search question should be chosen; what kind of investigation one

has to become engaged in; what kind of explanation one is looking

for; what kinds of categories will be permitted in theories; what

VC The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 184

Science and Public Policy, 43(2), 2016, 184–191

doi: 10.1093/scipol/scv029

Advance Access Publication Date: 16 June 2015

Article

kinds of empirical data one is going to seek out; how, when and to

whom one should present the findings.

The answers that are given to these questions determine what

particular scientific knowledge there will be. They determine on

which particular topics the subjects can provide any advice, of what

particular robustness it can be, what options for action will come to

light, and with what force it will be inserted into the policy arena.

Because of the impact of these choices scientific advisors can be

called agenda setters.

An array of different answers can be found in the literature as to

the question how researchers should deal with these issues. A well-

known answer is that of Max Weber (1958 (original 1918)).

Especially in Science as a Vocation, Weber spells out what the role

and position of the scientist towards politics and policy-making

should be. He maintains that the scientist can and should point out

inconvenient facts to politicians. What inconvenient facts are dis-

covered in the research depends on what the scientists decide is

worth knowing. The latter decisions inevitably follow the value pre-

suppositions of a particular scientific discipline (Weber 1958 (ori-

ginal 1918): 143, 146–8, 156).1

Wildavksy (1979) also answers the question by focusing on the

role of the individual scientist. He claims that the researcher should

not merely provide inconvenient facts in order to stimulate policy-

makers to re-evaluate their plans. According to Wildavsky scientific

advisors should themselves formulate (‘create’) the policy problems

that have to be solved. They should present policies to problems that

can be solved. Moreover, the expert advisors should use all their

rhetorical skills to persuade policy-makers to accept and implement

their feasible policies.

Others, however, argue that the questions mentioned should not

be answered soley by individual scientists. They maintain that scien-

tists do not have any privileged knowledge on these questions.

Moreover, leaving those agenda setting answers to scientists makes

them an elite, which is unwarranted in a democracy. These commen-

tators opt for the involvement of citizens or laypeople in the deci-

sions. On the particular form of citizens’ involvement and on the

issues on citizens should have a say, these authors differ, however

(Brown 2009; Catlaw 2007; Collins and Evans 2007; Flyvbjerg

2001; Kitcher 2001, 2011; Sarewitz 2011; Wagenaar 2011).

Some agree on the downsides of decision-making by scientists.

They opt, however, for forms of institutional design. To guarantee

that decisions are taken properly they turn to measures such as: peer

review in its different forms, orchestrated competition, regulation

and legal liability (see for instance the contributions in Lentsch and

Weingart (2011), Jasanoff (2011), Owens (2011), Rogers (2011),

De Wit (2011) and Stucke (2011) on peer review; Wagner (2011) on

competition; and the contributions by Den Butter (2011), Podger

(2011) and Van der Sluijs et al. (2011) on regulation).

2.1 Summing up This section started with an overview of the choices that scientific

advisors (or designers of advisory bodies and processes for that mat-

ter) have to make. A broad array of answers can be found to the

question: how should these choices be made? Some maintain that in-

dividual scientists should decide. Others opt for properly designed

advisory bodies and others, again, present lay involvement as the

proper approach.

These contradictory advocates for particular choices all agree on

one point. They all opt for a solution that is suitable for all occa-

sions. This observation leads to the suggestion that for more guid-

ance we may have to go beyond general answers. The different

answers found may all have their value, yet each has a particular

context. In recent publications an approach presents itself that seems

promising in this respect. This issue context approach offers a cat-

egorization of policy problems and shows how each type of problem

is properly dealt with.2

3. The issue context approach

Examples of the issue context approach can be found in studies that

otherwise differ in many respects. Bijker et al. (2009) present an

issue context framework to elaborate on an ethnographic case study

of a particular scientific advisory body. Pielke (2007) presents it as

part of a compelling advocacy for context sensitivity in scientific ad-

vice, illustrated with many American and also international issues.

Hoppe (2010) presents a thorough scientific study that builds on a

large set of references and that systematically connects the issue con-

text with many other aspects like epistemology, culture, participa-

tory design etc. Yet, the similarities between these studies warrant

the claim that they are all examples of one approach. In this section

the basic characteristics of this approach will be spelt out.

Elements of the issue context approach can also be found in ear-

lier writers. The idea of problems and problem contexts that de-

mand a certain approach can be recognized in Dunn’s analysis of

what he called the ‘wrong problem problem’ and the ‘principle of

methodological congruence’ (Dun 1988). The idea of a problem cat-

egorization along the dimensions of fact and consensus can be found

in Ezrahi (1980). A similar grid approach has been developed by

Douglas and Wildavsky (1983). None of these earlier writers, how-

ever, bring the different elements together into one approach.

3.1 Central aim The aim of the issue context approach is to improve the connection

between scientific research and policy-making. It intends to offer a

rational way to discern types of problems and to guide scientific ad-

visors and designers of advisory agencies in their decision-making.

By doing so it claims to offer better solutions in cognitive terms and

in terms of acceptability to stakeholders. In Hoppe’s words this

approach leads to the:

. . . maximization of the intelligence of democracy in the govern-

ance of problems. (Hoppe 2010: 249)

The advocates present their approach as an answer to new societal

circumstances. On the one hand, modern societies encounter ever

more complex problems and policy-makers turn to scientists for so-

lutions. On the other hand, there is a trend of declining civic trust in

science and scientists. These developments call for a reorientation of

the relation between science and policy-making and therefore on sci-

entific advisors and on nexus or boundary organizations in this field.

Being sensitive to issue contexts will help to choose better responses

and strengthen the trust in science (Pielke 2007: 18, 20–1; Bijker

et al. 2009: 157, 163–5; Hoppe 2010: 5–6, 42–3).

3.2 Typology of problems A basic element of the issue context approach is the distinction be-

tween different types of policy problems. Bijker et al. (2009: 159–

61) follow Renn (2005) in his distinction between four risk problem

characterizations: simple, complex, uncertain and ambiguous. This

distinction seems to express an increasing level of risk, yet it actually

expresses two different dimensions. Simple, complex and uncertain

all refer to the level of available knowledge in a particular situation.

Science and Public Policy, 2016, Vol. 43, No. 2 185

Ambiguous, however, brings a new element into play: situations in

which there is no agreement about shared values.

Pielke (2007) builds his argument on a number of extreme types

of politics or decision-making: tornado politics, abortion politics

and the gunman case. Tornado politics exemplifies the policy situ-

ation in which there is an evident consensus on the aim or values

(safety, survival for all people in this building) of the policy. In cases

of abortion politics there is absolutely no agreement on values and

aims. In tornado politics new information helps to solve the policy

issue (Information about whether a tornado is approaching is avail-

able. If a tornado approaches we decide to go to the shelter to-

gether.). In abortion politics new information, however reliable,

does not help to solve the policy issue. In the gunman case a pilot is

quite certain about his aims (avoiding an air collision with another

aircraft), yet he is provided with contradictory information (Pielke

2007). Here again two dimensions can be discerned: availability of

(certain) knowledge and clearness or consensus on values and aims.

Hoppe is clearest in distinguishing between two dimensions. He

presents a typology of four types of problems that result from the

combination of two dimensions (see Fig. 1).

The problem types used by Bijker et al. and Pielke can be situated

within this grid. With his dominating examples of tornado politics

and abortion politics Pielke focuses on structured and unstructured

problems (with the gunman as moderately (goal) structured). The

simple and complex problems used by Bijker et al. differ on the cer-

tainty of knowledge but involve high agreement on values.

Uncertain and ambiguous problems combine low knowledge cer-

tainty with moderate, or even little, agreement on values.

3.3 Different problems, different roles for scientific

advisors The authors in the issue context approach also distinguish between

different types of scientific policy advisors. The advisor types are

then connected to problem types.

Pielke presents four types: the pure scientist, the science arbiter,

the issue advocate and the honest knowledge broker. In the case of

tornado politics Pielke favors the science arbiter or the pure scientist.

The latter simply offers fundamental information without any inter-

est in the policy-making process. The science arbiter offers the

particular information that the policy-maker, given his current aims,

needs and understands. According to Pielke, in abortion politics the

role of the issue advocate and that of the honest broker are appropri-

ate. The issue advocate explicitly favors a particular aim or value

and offers the information that supports it. The honest broker of

policy alternatives:

. . . expands, or at least clarifies, the scope of choice for decision

making in a way that allows for the decision-maker to reduce

choice based on his or her own preferences and values.

This latter role can best be performed by a multi-disciplinary collec-

tion of experts (Pielke 2007: 1–3).

Bijker et al. opt for in-house advisors in the case of simple problems.

If problems are more complex in terms of knowledge, policy-makers

should turn to external experts. In the case of uncertain risk problems,

an advisory body that includes representatives of stakeholders is prefer-

able. When, for instance, there is no full scientific certainty on the toxic

risks of certain materials, input from industrial partners and labor

unions should have a role in determining what use is ‘as low as reason-

ably achievable’). Finally, ambiguous cases demand deliberative ar-

rangements with citizens more generally. Issues of, for instance, human

enhancement with nano-technologies or xeno-transplants should be the

subject of public debate (Bijker et al. 2009: 162–5).

In their characterization of different advisory roles Pielke and

Bijker et al. simply combine different types of elements: proper

orientation (detached or involved), organizational design, level of

citizen involvement etc. Hoppe discusses all of these elements separ-

ately and then combines them in his chapter on styles of advice.

Structured problems demand a rational style of advice. This can

take the form of cost–benefit analyses, audits, program evaluations,

or, more generally, evidence-based policy. Unstructured problems

need interactive, participatory or argumentative styles. The argu-

mentative style is for the ‘policy philosopher’ who makes people

aware of other possible worldviews and their implications. The

interactive style is that of the mediator who, through continuous

interaction, guides learning processes. In the participatory style the

advisor tries to resist elitist bias by including other societal values

and interests. Moderately structured problems in which there is con-

sensus on goals should be dealt with in client advice or process

Agreement on

norms and values

Certainty on

knowledge

Low High

Low Unstructured problems Moderately structured problems (goals)

High Moderately structured problems (means)

Structured problems

Figure 1. Typology of problems (Hoppe 2010: 16, 73)

186 Science and Public Policy, 2016, Vol. 43, No. 2

management styles. Prominent in both is offering strategic advice

that enables clients to achieve their goals. In the last type of prob-

lems, the moderately structured ones in which information on means

is clear, the argumentative and the interactive style are appropriate

(Hoppe 2010: 188–94). The true high-achieving advisor is familiar

with each problem type and chooses and employs the proper style

accordingly. Hoppe (2010: 116) refers to such a person as ‘Hermes’.

3.4 Concern for wrong problem problems The advocacy for linking particular advisory roles to problem types

is mirrored in a concern to avoid what Dunn (1998) has called

wrong problem problems. Pielke emphasizes that dealing with tor-

nado situations and abortion situations in the wrong way is not very

helpful in solving the problem. In a really effective democracy all the

different types of advisors have their due. Only then can scientists:

. . . contribute productively to effective decision-making and sus-

taining the long-term viability of the scientific enterprise. (Pielke

2007: 18, 21)

Hoppe claims that currently there often is a mismatch between the

problem perceptions of policy-makers and their advisors on the one

hand and the perceptions of large parts of society on the other:

Better governance implies political sensitivity to different types of

problems; and more and better reflexive problem structuring

through better institutional, interactive and political designs for

public debate and political choice. (Hoppe 2010: 6)

One type of mismatch is most prominent in the issue context ap-

proach. Pielke is most concerned that abortion-type problems (huge

disagreement on values) are being addressed as if they were of the

technical tornado type. When scientific advisors present their know-

ledge without regard to the existing diversity of values (thus as a

pure scientist or as a science arbiter, when they should be issue advo-

cates or knowledge brokers) they are actually operating as stealth

issue advocates. They do not enlarge but reduce the scope of choice

available to decision-makers, as is the defining characteristic of issue

advocates. Yet, they do not do so openly and may not even be aware

of doing so. According to Pielke (2007: 94–5), stealth issue advocacy

poses threats to the scientific enterprise as it tends to allow the pub-

lic to believe that scientific findings are simply an extension of a sci-

entist’s political convictions. He presents the example of scientific

contributions of members of the International Panel on Climate

Change. Insensitivity to value plurality led to accusations of politics

dressed up as science. The advisory role that Pielke (2010: 141)

favors most is that of the honest broker of policy alternatives that

clarifies and expands the range of options.

Hoppe shows concern for the same issue. He discerns a tendency

in policy-making and advice to deal with unstructured problems as

if they were structured. Advisors and policy-makers hold onto cer-

tain organizational cultures, issue networks, types of analyses and

forms of civic participation although they are inappropriate for such

problems. In fact, they try to domesticate unstructured problems.

Hoppe maintains that this particular mismatch explains the dimin-

ishing support for policies and politics. Policy-making and advisory

proccesses should take the unstructured character of many problems

into account and therefore turn to more participatory and delibera-

tive approaches. This will lead:

. . . to [a] rebalance [of] democratic politics towards the pole of

reason, deliberation and learning, and away from the pole of

power backed by authority and potential coercion. (Hoppe 2010:

251 also: 20, 75, 215, 220, 223)

Collaborative hybrid forums or platforms and knowledge centers

are powerful in dealing with unstructured problems. They are the

hothouses for innovative policy ideas (Hoppe 2010: 231).

Bijker et al. (2009: 163) have a somewhat different concern on

this point. They witness a growing demand for scientific advisory

bodies to give advice on, what Hoppe calls, unstructured problems.

In such cases they think public debate is appropriate. They opt, how-

ever, for arrangements in which civic deliberation is kept separate

from the work of advisors and advisory bodies. While Hoppe and

Pielke argue for the active involvement of scientific advisors in pub-

lic deliberation on values and aims, Bijker et al. consider such an ap-

proach to be problematic. They fear that it would undermine the

authority of the advisory body.

3.5 Institutional design This latter remark already points to the field of institutional design.

All three advocates of an issue context approach provide guidelines

for the organization of advisory processes and advisory bodies. At

least three recommendations can be formulated.

First, it is maintained that the more unstructured the problem,

the more civic involvement in realizing advice is called for.

Particularly in these cases, collaborative hybrid arrangements are

appropriate. Citizens can thus counterbalance elitist tendencies in

research and advice. Furthermore, scientific advisors can clarify

diverging valuations and bring together opposing social groups

(Hoppe 2010: 220ff, 231; Bijker et al. 2009: 155, 164–5; Pielke

2007: 94).

Second, the more unstructured the problem, the more need there

is for a multi-disciplinary community of scientific advisors. Pielke

(2007: 151, 154) maintains that his favored style of honest broker-

ing can best be performed by an interdisciplinary team of experts.

According to Hoppe (2010: 231) the broad platforms and know-

ledge centers are the contemporary hothouses of innovative policy

ideas. Bijker et al. (2009: Chap. 3, 163) emphasize the advantages of

the multi-disciplinary composition of the advisory body they

studied.

The final recommendation follows the claim that the more un-

structured the problem, the more the involvement of scientific ad-

visors in the process of goal-finding is appropriate. The advocates of

the issue context approach all opt for the involvement of scientists in

the process of public value deliberation—be it in different forms and

degrees. Pielke (2007) presents explicit issue advocacy as a perfectly

valid strategy in abortion-type problems. Hoppe (2010: 231) sup-

ports a creative and critical role of scientists in policy debates and

also as active moderators of mediation. Bijker et al. (2009: 161,

163) seem to have reservations on this point. Yet, they also refer to

and endorse the actions of the advisory body they studied where it

was clearly trying to influence the acceptance of certain advice.3

3.6 Summing up The advocates for an issue context approach differ on certain points,

but the main lines of their proposals and their concerns are quite

similar. They all opt for a distinction in problem types following di-

mensions of certainty of knowledge and agreement on aims. Types

or styles of scientific advice should follow problem type. In broad

terms: evidence-based research and a detached attitude are appropri-

ate for the more structured problems and a critical and creative

Science and Public Policy, 2016, Vol. 43, No. 2 187

approach combined with broad civic deliberation for the unstruc-

tured problems.

The issue context approach (in these advocacies) might not pre-

sent clear answers to all the value issues that were distinguished in

Section 2. Yet, it promises to offer a useful frame of reference for

many judgments that scientific advisors or designers of advisory

bodies have to make.

4. Assessing the issue context approach

This section first concentrates on the possibility of categorizing

problems according to the dimensions of the issue context. It then

discusses the connection between problem types and the roles of sci-

entific advisors.

4.1 On categorizing problems A problem that one immediately encounters if one tries to apply the

issue context approach is that the authors do not provide clear indi-

cators of the two dimensions into which problems are to be catego-

rized. Pielke (2007) presents us the examples of tornado politics and

abortion politics, but he does not make explicit when, exactly, a

societal phenomenon is an issue on which there is (low or high)

agreement on values or on knowledge. Neither do Bijker et al.

(2009). Even in the broad ranging argument of Hoppe (2010) the di-

mensions are presented as if they were self-evident.

Furthermore, the authors are not particularly clear about the

meaning of the dimensions themselves. For instance, in Hoppe

(2010: 16) the dimension: ‘level of agreement on norms and values

at stake’ might be interpreted in different ways. It might refer to the

level of consensus on what particular values and aims are of rele-

vance or of importance (for at least some participants). Yet, it might

also be interpreted as the level of consensus among all on (the prior-

ity of) the aims and values. In the first understanding there is only

agreement on the list of aims that are valued by some. In the second

understanding there is a substantial consensus on the aims. In Bijker

et al. (2009: 162–3) these two particular understandings seem to be

distinguished as two different categories. Each type of agreement de-

mands a different type of approach. As to uncertainty on knowledge

Pielke distinguishes between uncertainty in terms of risk probabil-

ities and uncertainty in terms of having any knowledge of conse-

quences (Pielke 2007: Chap. 5). The other advocates of the issue

approach are less explicit on this point. In all, we have to conclude

that the authors are not very clear on the central terms in their

approach.

On the level of indicators for the two dimensions (for the mo-

ment neglecting the ambiguities mentioned above) we come across

another set of difficulties. First, the question arises whether the re-

searchers are talking about consensus and disagreement in political

arenas or in a broader societal context. An issue might be fiercely

debated in a particular public arena (e.g. parliament) without having

any real resonance in society. The reverse might also be the

case: some societal issues might not reach the public stage

(cf. Gusfield 1981: 5). To put the same point differently: whose

understandings of the issue have to be taken into account in deter-

mining the character of the problem? Who exactly are ‘those

involved’ (Hoppe 2010: 64)?

In some cases, particular actors may have an interest in trying to

manipulate the (public) understanding on an issue. Tobacco compa-

nies, for instance, might try to create the impression of scientific dis-

agreement on the harms of smoking. Collectively, politicians might

have reasons to neglect a particular, socially controversial, issue.

Also from the side of scientists there can be interest-based reasons

for trying to exaggerate or to minimize the actual (dis)agreement on

specific knowledge (Ezhari 1980). In general, many actors may have

reasons to diminish or exaggerate levels of agreement. This phenom-

enon makes it highly complicated to pinpoint the ‘real’ level of

agreement.

Yet, even if it were clear what the two dimensions exactly refer

to and by which indicators we could measure them, there are further

complications for the issue context approach. First, understandings

of problems, and therefore (dis)agreement on relevant values and

knowledge, are not fixed over time. Issues in, for instance, environ-

ment, medicine and healthcare, food technology, migration or secur-

ity seem to change rapidly. After periods of (seeming) agreement

controversies arise, and vice versa. The level of agreement can also

differ across different groups, societies and cultures at any particular

moment (Brown 2008; Jasanoff 2008).

On a more abstract level, questions also arise as to the two-

dimensions characterization of issues. Are the knowledge dimension

and the value dimension really independent? This is not the place to

elaborate on the fact–value dichotomy and its philosophical

problems (see Douglas (2009) for a comment on naı̈ve fact–value

distinctions). Yet, on a more practical level the dichotomy, as it is

used in the issue context approach, is also questionable. The logic of

this approach has it that developments in the one dimension are in-

dependent of the other. Yet, the presentation of new scientific know-

ledge may very well have an impact on the value or aim side of

problems. New findings in medicine might influence people’s ideas

on, and valuation of, what it means to have a life that is worth liv-

ing. Developments in computer and surveillance technology, to give

yet another example, may cause people to re-evaluate their under-

standings of freedom and privacy.

The conclusion, so far, must be that the advocates of the issue

context approach are not very helpful in objectively determining the

type of problem in a particular case. Moreover, such an objective

problem determination on the two dimensions is problematic.

In some instances the advocates seem to recognize these prob-

lems. For instance, Bijker et al. (2009: 162) recognize the diversity

and fluidity of problems over time and cultures. Hoppe (2010: 72,

245) notes that politicians create problem definitions that fit their

interests best (‘acceptability heuristic’). More generally, both present

their approach as social constructivist (Bijker 2009: 3–5, 139;

Hoppe 2010: 50, 64), thereby implying that objective renderings of

social reality (including problem definitions) cannot be given. These

text elements, however, do not fit well with the overall argument of

the advocates. The very idea of wrong problem problems demands

that problems have their objective characteristics. It demands that

the real issue context in terms of certainty and value agreement can

be determined.

4.2 On problem typology as a principle for design Having discussed the objective determination of issues, we can now

turn to a second aspect: the proper attitude for scientists towards

each of the types of issues. On a closer inspection, shortcomings and

problems also occur here.

The advocates of the issue context approach connect, as pointed

out above, issue types and appropriate roles and attitudes of scien-

tists. They neither build, nor do they suggest that, their claims have

a grounding in empirical research. Their claims are developed and

displayed in particular examples. Pielke builds his case largely on

188 Science and Public Policy, 2016, Vol. 43, No. 2

the two extreme, and somewhat artificial, cases of tornado politics

and abortion politics. The claim that in the first case a pure scientist

and in the second a honest broker are the appropriate roles might

have its intuitive appeal—it is solely this intuition that has to carry

the whole weight of Pielke’s argument. Bijker et al. build their case

on an ethnographic study of one advisory body. Hoppe refers to a

huge body of literature. However, most of his material serves to sup-

port the many distinctions he makes, not the central claim that con-

necting problem types and the roles of scientists leads to better

results.4

Thus, the advocates develop and illustrate their claims by dis-

playing exemplary cases, real or imagined. These examples cannot,

however, furnish much support to the claims. One might ask, for ex-

ample why we should (always) deal with cases of high normative

disagreement in the ways that the advocates want us to. Why in

such cases should new answers be developed in deliberative settings

with citizen groups (Hoppe’s ‘mini-republics’)? Why should a scien-

tist in these cases broaden the issue even further as a honest broker

(Pielke)? It can be imagined that in some cases the ongoing debate

on a public issue is best solved by allowing an authoritative body to

present an answer. Solving a problem might also demand that issues

are disentangled (or that the problem is cut into pieces) or com-

bined. In Wildavky’s words:

. . . sometimes problems have to be created that can be solved

(Wildavsky 1979: 17).5

To take on a further point: why should a scientist respect some exist-

ing (public) consensus on an issue and limit himself or herself to pro-

viding mere instrumental scientific input? Why would it be wrong if

a scientist presented findings that shattered the (alleged) public con-

sensus? These findings might, for instance, show that because of

some shared understandings (of policies or states of affairs) negative

consequences for a particular group (e.g. women, or future gener-

ations) have been neglected. Scientists might point out that certain

problem definitions and values are dominating a discussion while

others might also be taken into account. (See for instance the ex-

ample of a dominating concern for state and market and a neglect of

civil society (Burawoy 2005).) In general terms, the issue context ap-

proach seems to have a strong conservative (or conventionalist) bias.

It is not clear why simply presenting new scientific findings (as a

pure scientist, as a reducer of scientific uncertainty, as an adept of

the rational style) cannot lead to people changing their minds on

multi-valued or value conflict issues. Even in Pielke’s case of abor-

tion it might very well be the case that new scientific findings (new

medical techniques) bring people to alter their views (cf. Brown

2008; Jasanoff 2008). New technology might change what we value

and what we value it for (this brings us back to the earlier comment

on the fact–value dichotomy).

A last critical point in this line revolves around the issue why sci-

entists should have a role in mediation (Hoppe) and brokering

(Pielke) in cases of value disagreement? Bijker et al. have their reser-

vations on this point. Indeed one wonders whether scientists are the

best agents for such a task. Do they (all) have the skills to perform

such tasks best? Is it by being an active mediator that scientists can

introduce their knowledge into the deliberation? None of the advo-

cates provides a convincing basis for a positive answer to these

questions.

A third issue that raises doubt about the advocates’ claims is

their disagreement on certain points. Here we can recall examples

given above. The authors differ on the issue of civic participation.

They also have different opinions on the appropriateness of

scientists acting as issue advocates. More generally, it might be con-

cluded that they reason from different understandings of what

makes a problem solution a good solution. Only Hoppe is explicit

on this point. His basic understanding of politics as a dialectical and

interrogative process (Hoppe 2010: 8) brings him close to a

Habermassian position, with great concern for dialogue and civic in-

volvement and finding shared problem frames. Pielke is much less

explicit. His view on politics, if we try an interpretation, seems to be

more Weberian, given its focus on strategies and advocacy. In con-

trast, Bijker et al. seem to steer a more elitist Aristotelian course.

Scientists in an advisory body should try to find the best answer for

a particular issue. In a process of internal deliberation they should

take all relevant issues into account, including the typical sensitiv-

ities of certain groups.

A last issue that has to be considered is whether the guidelines of

the issue context advocates are likely to contribute to (a strengthen-

ing of) the legitimacy of science in society. One aspect already ad-

dressed above is whether the approach is likely to solve problems

better. Another point is whether it contributes to social trust in sci-

ence as a politically neutral or autonomous institution. All advocates

considered here underline the importance of this issue. Yet, only

Bijker et al. (2009: 165) explicitly take it into account when discuss-

ing the proper roles of scientists. As to cases of strong value dis-

agreement they express their concern that advisory bodies might

become over-involved politically. Pielke and Hoppe, to the contrary,

also favor scientific involvement in these types of cases (issue advo-

cacy and brokering in Pielke, involvement in mediation and deliber-

ation in Hoppe). Possible negative consequences—a societal image

of scientists as agents of special interest—are not considered. The

latter does, of course, not mean that truth is on the side of Bijker

et al. here. One can imagine cases where it seems quite reasonable

and appropriate for individual scientists or advisory bodies to pre-

sent advice that emphasizes a particular value or goal. Researchers

or expert advisors might have very strong indications that a particu-

lar wrong or harm is likely to occur: a wrong or harm that (in public

discussion) has not yet been recognized as such. Thus, the appropri-

ateness of issue advocacy seems to rely not on public agreement or

disagreement but on the demands of responsibility on the part of the

scientists and experts. The more certainty they have of a possible

wrong or harm that is likely to occur and the less likely it is that

others will recognize it, the larger is their responsibility to try to

place it on the political agenda (cf. Thompson 1983).

In sum the conclusion must be that it is doubtful that the issue

context approach will bring what advocates expect of it: that is a

better solution of problems and societal legitimacy for scientific

practice.

5. Conclusions

This paper has concentrated on the issue context approach because

of its promising claim to connect problem types with advisory and

scientific strategies. If valid this approach could provide guidelines

for individual researchers and rules for the institutional design of ad-

visory bodies.

Unfortunately, the issue context approach cannot live up to this

promise. First, its advocates do not provide us with the tools for an

objective determination of problems. It is also unlikely that such ob-

jective indicators can be given. Furthermore, the guidelines for ad-

visory rules and institutional design are not very convincing. The

empirical claims rest on a few examples and can easily be challenged

by counter examples.

Science and Public Policy, 2016, Vol. 43, No. 2 189

The shortcomings of the issue context approach are partly of an

empirical kind (no clear indicators, no firm evidence etc.). Yet it also

suffers from insufficient reflection on its own normative presuppos-

itions. Such a reflective mode would involve asking questions like

‘What is to be taken as a (good) solution to a problem and why?’

Bijker et al. and Pielke do not address such issues. (Although one

might discern an adherence in them to Aristotelian or Weberian phi-

losophies, as mentioned above). Hoppe (2010: 7–8) is more explicit.

He starts from an understanding of human nature that goes beyond

that of a homo cogitans—he opts for the more holistic homo respon-

dens (mentioning Arendt and Ricoeur among others as inspiration).

Yet, it remains unclear how this understanding of human nature

leads to his particular claims on approaches to be chosen when en-

countering particular types of problems. In fact, his approach has a

conservative streak—policy definitions should follow common

understandings and agreements—that does not do justice to

Arendt’s analyses.

Issue context approaches, especially in the unreflective form used

by Bijker et al. and Pielke, basically intend to provide answers to

value issues by relying on scientific findings: if you encounter such a

problem, you should do this.

Certainly, the studies which we have examined are of value.

They show the types of value decisions that scientific advisors have

to make. Second, they make us aware of the ways in which problem

settings can differ and that it is highly unlikely that a strategy that

works on some issues will have the same effect on others. Third, the

advocates offer overviews of possible strategies and value orienta-

tions. In all, they help to make us more aware of what is at stake,

and what the difficulties are in scientific advice. However, they can-

not provide a shortcut through the difficult field of value choices in

scientific advice.

A basic aim of public administration as a discipline is making

policy and government more rational, giving it a firmer grounding

in empirical research and systematic analysis. These motives are

laudable, yet there are limits to this ambition. This assessment of the

issue context approach concerns such limits. Scientific advisors have

to make choices. And the answer to the question of what choices are

proper cannot be deduced from scientific findings—this was one of

the lessons learnt from Max Weber. Advocates of the issue context

approach and others who nevertheless try this road encounter the

problems that were pointed out in this paper. In as far as scientists

still hold onto such ideas they risk leading public administration to

scientism and managerialism: claiming scientific authority on issues

where science cannot have authority.

The issue context approach cannot provide the shortcut through

value choices that it promises. Inevitably, real value decisions have

to be made by scientific advisors (or designers of advisory bodies).

What can really be of help for scientists and advisors is the philo-

sophical elaboration of relevant values and guiding principles. There

already are many examples of such elaborations. For example, Elliot

(2011: Chap. 3) has argued for taking the ‘informed consent’ of citi-

zens and their representatives as a guiding principle in decision-mak-

ing on science and policy. Others have emphasized the importance

of the harm principle in these matters and evaluated the role that a

cautionary principle might have (Sunstein 2005; Luján and Todt

2012; Holbrook and Briggle 2014). A third example is the ethical re-

search that has been done into types of responsibility and the scope

of the responsibility of policy advisors (Thompson 1983).

Philosophical elaborations of the values and possible guidelines and

principles for scientific advisors’ choices like these are of great rele-

vance. They can help to structure and strengthen individual

judgment. In order to have such an impact it must find its way into

academic practice and university curricula. This, however, brings us

to a theme that goes beyond this paper.

Notes 1. For an overview of Weber’s position and arguments, including

other works than Science as a Vocation, see Ciaffa (1998).

2. The term ‘issue context’ is used in Letsch and Weingart (2011:

10 note 8). In this paper three particular examples of this ap-

proach will be examined. The approach, more or less de-

veloped, can also be found in others, however.

3. The active and judgmental role of the advisory body is some-

what veiled by Bijker et al. They use terms like the ‘balancing

of interests’ (Bijker et al. 2010: 163, 165) suggesting a strictly

neutral stand. Yet, a balancing of interests implies a particular

pre-understanding on the side of the scientific advisors of what

are the relevant interests and what may count as a proper

balance.

4. For example, Hoppe’s central chapter on scientific roles builds

to a large extent on a paper by Mayer et al. (2004) that itself

only offers a framework and no empirical analysis (Hoppe

2010: Chap. 7, esp. 184ff.)

5. To make the same point differently: the authors do not make

explicit what they mean by ‘solving’ a problem and why we

should accept this interpretation.

References

Bijker, W. E., Bal, R. and Hendriks, R. (2009) The Paradox of Scientific

Authority. The Role of Scientific Advice. Cambridge, MA: MIT Press.

Brown, M. B. (2008) ‘Review of R. A. Pielke: The Honest Broker’. Minerva,

46: 485–9.

—— (2009) Science in Democracy. Expertise, Institutions and Representation.

Cambridge, MA: MIT Press.

Burawoy, M. (2005) ‘For public sociology’. American Sociological Review,

70: 4–28.

Catlaw, T. J. (2007) Fabricating the People. Politics and Administration in the

Biopolitical State. Tuscaloosa, AL: University of Alabama Press.

Ciaffa, J. A. (1998) Max Weber and the Problems of Value-free Social Science.

A Critical Examination of the Werturteilsstreit. London: Associated

University Press.

Collins, H. and Evans, R. (2007) Rethinking Expertise. Chicago, IL:

University of Chicago Press.

Den Butter, F. (2011) ‘The industrial organization of economic policy prepar-

ation in the Netherlands’ in The Politics of Scientific Advice. Institutional

Design for Quality Assurance, J. Letsch and P. Weingart (eds), pp. 177–

214. Cambridge, UK: CUP.

De Wit, B. (2011) ‘RMNO and quality control of scientific advice to policy’ in

The Politics of Scientific Advice. Institutional Design for Quality Assurance,

J. Letsch and P. Weingart (eds), pp. 139–56. Cambridge, UK: CUP.

Douglas, H. (2009) Science, Policy and the Value-free Ideal. Pittsburgh, PA:

University of Pittsburgh Press.

—— and Wildavsky, A. (1983) Risk and Culture. Berkeley, CA: University of

California Press.

Dun, W. (1988) ‘Methods of the second type: Coping with the wilderness of

conventional policy analysis’. Policy Studies Review, 7: 720–37.

Elliott, K. (2011) Is a Little Pollution Good for You? Incorporating Societal

Values in Environmental Research. Oxford, UK: OUP.

Ezrahi, Y. (1980) ‘Utopian and pragmatic rationalism: The political context of

scientific advice’. Minerva, 18: 111–31.

Flyvbjerg, B. (2001) Making Social Science Matter. Why Social Inquiry Fails

and How it Can Succeed Again: Cambridge, UK: CUP.

Gusfield, J. R. (1981) Drinking–Driving and the Symbolic Order. The Culture

of Public Problems. Chicago, IL: University of Chicago Press.

190 Science and Public Policy, 2016, Vol. 43, No. 2

Holbrook, J. and Briggle, A. (2014) ‘Knowledge kills action – Why principles

should play a limited role in policy making’. Journal of Responsible

Innovation, 1: 51–66.

Hoppe, R. (2010) The Governance of Problems. Puzzling, Powering and

Participation. Bristol, UK: Policy Press.

Jasanoff, S. (2008) ‘Speaking honestly to power. Review of the Honest Broker:

Making Sense of Science in Policy and Politics, by R. A. Pielke’. American

Scientist, 96(3): 240–3.

—— (2011) ‘Quality control and peer review in advisory science’ in

The Politics of Scientific Advice. Institutional Design for Quality

Assurance, J. Letsch and P. Weingart (eds), pp. 19–35. Cambridge, UK:

CUP.

Kitcher, Ph. (2001) Science, Truth and Democracy. Oxford, UK: OUP.

—— (2011) Science in a Democratic Society. Amherst, NY: Prometheus

Books.

Lacey, H. (2005) Is Science Value Free. Values and Scientific Understanding.

London: Routledge.

Letsch, J. and Weingart, P. (eds) (2011) The Politics of Scientific Advice.

Institutional Design for Quality Assurance. Cambridge, UK: CUP.

Longino, H. (2002) The Fate of Knowledge. Princeton, NJ: Princeton

University Press.

—— (2004) ‘Science and the common good: Thoughts on Philip Kitcher’s

Science, Truth, and Democracy’. Philosophy of Science, 69: 560–8.

Luján, J. and Todt, O. (2012) ‘Precaution: A taxonomy’. Social Studies of

Science, 42: 43–57.

Mayer, I.S., Van Daalen, C.E., Bots, P.W.G. (2004) ‘Perspectives on Policy

Analysis. A Framework for Understanding and Design’. International

Journal of Technology, Policy and Management, 4/2: 169–91.

Owens, S. (2011) ‘Knowledge, advice and influence: The role of the UK Royal

Commission on Environment Pollution 1979–2009’ in The Politics of

Scientific Advice. Institutional Design for Quality Assurance, J. Letsch and

P. Weingart (eds), pp. 102–13. Cambridge, UK: CUP.

Pielke, R. (2007) The Honest Broker. Making Sense of Science in Policy and

Politics. Cambridge, UK: CUP.

Podger, G. (2011) ‘Quality control and the link between science and regulation

from a national and EU administrator’s perpective’ in The Politics of

Scientific Advice. Institutional Design for Quality Assurance, J. Letsch and

P. Weingart (eds), pp. 229–37. Cambridge, UK: CUP.

Renn, O. (2005) White Paper on Risk Governance: Towards an Integrative

Approach. Lausanne, Switzerland: International Risk Governance Council.

Rogers, M. D. (2011) ‘The European Commission and the collection and use

of science and technology advice’ in The Politics of Scientific Advice.

Institutional Design for Quality Assurance, J. Letsch and P. Weingart (eds),

pp. 115–36. Cambridge, UK: CUP.

Sarewitz, D. (2011) ‘Looking for quality in all the wrong places, or: The

technological origins of quality in science policy advice’ in The Politics of

Scientific Advice. Institutional Design for Quality Assurance, J. Letsch and

P. Weingart (eds), pp. 54–70. Cambridge, UK: CUP.

Stucke, A. (2011) ‘Quality assurance through procedures – policy advice by

the German Science Council’ in The Politics of Scientific Advice.

Institutional Design for Quality Assurance, J. Letsch and P. Weingart (eds),

pp. 157–73. Cambridge, UK: CUP.

Sunstein, C. (2005) Laws of Fear: Beyond the Precautionary Principle.

Cambridge, UK: CUP.

Thompson, D. F. (1983) ‘Ascribing responsibility to advisors in government’.

Ethics, 93: 546–60.

Van der Sluijs, J. P., Petersen, A. and Funtowitz, S. (2011) ‘Reflective

approaches to uncertainty assessment and communication’ in The Politics

of Scientific Advice. Institutional Design for Quality Assurance, J. Letsch

and P. Weingart (eds), pp. 259–69. Cambridge, UK: CUP.

Wagenaar, H. (2011) Meaning in Action. Interpretation and dialogue in

Policy Analysis. New York: Sharpe.

Wagner, G. (2011) ‘Quality control for the leading institutes of economic re-

search in Germany: Promoting quality within and competition between the

institutes’ in The Politics of Scientific Advice. Institutional Design for

Quality Assurance, J. Letsch and P. Weingart (eds), pp. 215–28.

Cambridge, UK: CUP.

Weber, M. (1958 (original 1918) ‘Science as a vocation’ in From Max Weber:

Essays in Sociology, H. H. Gerth and C. Wright Mills (eds), pp. 129–56.

London: Routledge and Kegan Paul.

Wildavsky, A. (1979) ‘Analysis as art’ in Speaking Truth to Power, pp. 1–19.

Piscataway, NJ: Transaction Publishers.

Science and Public Policy, 2016, Vol. 43, No. 2 191

Copyright of Science & Public Policy (SPP) is the property of Oxford University Press / USA and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.