Garce, A., Uña G. (2010) Think Tanks and Public Policies in Latin America,

7/30/2019 Garce, A., Ua G. (2010) Think Tanks and Public Policies in Latin America,

1/319


2/319

Think Tanks and

Public Policies

in Latin America

Edited by Adolfo Garc and Gerardo Ua

This publication reports on a research project financed by Canada's

International Development Research Centre (www.idrc.ca).


3/319

Garc, Adolfo and Gerardo Ua: Think Tanks and Public

Policies in Latin America, Fundacin Siena and CIPPEC,

Buenos Aires, Argentina, 2010.

Graphic design: sarapaoletti.blogspot.comTranslation: Patricia Draper and Debbie Gonzlez

Canada (Introduction)

Edition: Luciano Strazza and Dolores Arrieta

Think Tanks and Public Policies in Latin America 2


4/319

INDEX

PART I. Methodological and Conceptual Aspects on the Relationship between Think

Tanks and Governments

ABELSON, D., Is Anybody Listening? Assessing the Influence of Think TanksPage 11 BAIER, G. y BAKVIS, H., Think Tanks and Political Parties in Canada: Competitors orCollaborators?Page 33 DATTA, A., JONES, N., MENDIZBAL, E., Think Tanks and the Rise of the KnowledgeEconomy Their Linkages with National Politics and External Donors.Page 46

PART II. Think Tanks in Latin America

BRAUN, M., CHUDNOVSKY, M., DUCOT, N. y WEYRAUCH, V., Far away fromThinktankland: Policy Research Institutes in Developing Countries.Page 74 BELLETTINI, O., The Role of Public Policy Centers in Public Reforms Implemented in LatinAmerica.Page 106 UA, G., LUPICA, C., STRAZZA, L., Think Tanks and Poverty in Latin America: The Roleof Thinkers in the Marketplace of Social Policies in Argentina, Chile and Mexico.

Page 127 LVAREZ RIVADULLA, J., MARKOFF, J. y MONTECINOS, V., The Trans American MarketAdvocacy Think Tank Movement.Page 172

PART III. Think Tanks in the Ro de la Plata

CAMOU, A., Knowledge Behind the Throne. Expert Intellectuals, Think Tanks and EconomicPolicies in Argentina: between the Austral Plan and the Convertibility CrisisPage 209

UA, G., Think Tanks in Argentina: Overcoming the Tension between Participation andPermanence.Page 238 GARC, A., A Narrow and Unstable Interface. Think Tanks and Political Parties in Uruguay.Page 271 GALLARDO, J., GARC, A. y RAVECCA, P., Think Tanks and Experts in the Frente AmpliosGovernmentPage 290



5/319

Introduction.

Think Tanks and Public Policies

in Latin America

BY ADOLFO GARC AND GERARDO UA

During the year 2005, the authors of this book agreed on promoting a workshop calledThink tanks and public policies: Global Dynamics and Local Specifications, in the context ofthe First International Forum on the Social Science Policy Nexus, organized by theManagement of Social Transformations Programme (MOST) that UNESCO was planning tocarry out in Argentina and Uruguay. Convincing the Forum organizers of the pertinence of ourproposal was not difficult. In fact, the area specialists admit that the complex connection

between social research and public policies has an increasing participation of this diversegroup of institutions so-called think tanks. After a long round of contacts, and with the supportof the Regional Office for Latin America and the Caribbean of the International DevelopmentResearch Center (IDRC) from Canada and the Argentinean office of the Konrad AdanauerFoundation, we finally achieved to confirm the participation of both national and internationalrenowned academics and experts, as well as politicians and directors of the most importantArgentine think tanks. The workshop took place on the 21st and 22nd of February, 2006, inBuenos Aires. As the different meetings took place and as the speakers presented their pointsof view, the idea of gathering their expositions in a book came up, so that politicians, special-

ists and citizens of the region could access to the most recent discussions about think tanksin the world and the region. We would like to thank, once again, the support of IDCR and theKonrad Adanauer Foundation for making this initiative real.

A great number of the articles published in this English edition were part of the 2007 Spanishedition of Think Tanks and Public Policies in Latin America. Both of them gather edited versionsof some of the articles presented in that Workshop. That is the case of the chapters written byDonald Abelson, Gerald Baier and Herman Bakvis, Antonio Camou, Adolfo Garc and GerardoUa. Together with these texts, the reader will also have access to a number of articles writtenin other contexts and times. The chapters signed by Orazio Bellettini, Miguel Braun, Mariana

Chudnovsky, Nicols Ducot and Vanesa Weyrauch belong to this last category. Plus, four newunpublished chapters have been specially added for this English version: the first of them waswritten by Ajoy Datta, Nicola Jones and Enrique Mendizbal, members of the OverseasDevelopment Institute (UK); the second, by Gerardo Ua, Carina Lupica and Luciano Strazza,members of Fundacin Siena (Argentina), was done with IDRCs support; the third one is thework of Mara Jos lvarez Rivadulla, John Markoff and Vernica Montecinos, from PittsburgUniversity (USA), and the fourth of them was developed by Javier Gallardo, Adolfo Garc andPaulo Ravecca, from Uruguay Republic University, whom also received IDRCs support.

As it may be seen, this book is the result of a great communication effort between diverse



6/319

groups, all of which are interested in the influence of social research in public policies. This

work brings together North American, Latin American and European academics. At the same

time, this book also gathers both think tanks consultants and managers, with their own prac-

tical perspectives. So far, and just as an example, the splendid theoretical approach of Donald

Abelson regarding think tanks had not yet been contacted to John Markoff and Vernica

Montecinos contribution to the study of the influence of economists on public policies. At thesame time, these two articles had not been contacted with ODIs Research and Policy in

Development (RAPID), a crucial program that contributes to the promotion of the interface

between research and policies. It is comforting to know that, from South America, from the Ro

de la Plata, we were able to build a broad, qualified and innovative network on such a central

topic for contemporary democracies.

These new articles explore important dimensions of the development and operations of

think tanks, such as the bonds with the institutional environment they are inserted in, the rela-

tionship with the donors, the exchange with similar organizations in developed countries and

the dynamics with political parties. Furthermore, the increasing participation that Latin

American think tanks are gaining in the social policies cycle is also studied. In this way, the new

articles enrich the diversity of the book and provide new perspectives of analysis upon critical

aspects that take part in these organizations institutional development and their ability to influ-

ence public policies. These new perspectives highlight the prominent role think tanks might

have in the social policies arena, in a context where there is consensus on fighting poverty as

one of the most central challenges of the region, challenge that will require active participation

from both the State and civil society.

* * *

So far, we have synthetically introduced the genesis of this book. Allow us, before we continue

to present its content, to explicit a few a priorihypotheses that we considered to be particularly

relevant. In the first place, as our reader has probably noticed, we have omitted to define precise-

ly what kind of organizations are referred to as think tanks in recent literature. As it will be shown

in the different chapters of the book, theres no agreement among specialists on the precise defi-

nition of this term. As a matter of fact, as the study cases grow, it becomes evident that a great

range of institutions are called think tanks. In the United States, for instance, it is common prac-tice to designate private organizations, financed by companies and specialized in public policies

analysis, as think tanks. In other countries, the task of improving public policies quality through

social research is done by governmental institutions or university centers. Like any other global

phenomenon, the expansion of think tanks manifests itself in a specific way in every local context.

Evidence provided by study cases and comparative research suggests that national and / or

regional traditions shape in different ways this kind of organizations.

Secondly, we would like to explain the way in which we understand the link between special-

ized knowledge and public policies or, in other words, between experts and politicians. It is impor-



7/319

tant to clarify this, since some recent political experiences in the region have given a negative con-notation to the participation of experts in the Government. We are convinced that the developmentof the nations require that Governments support their work, amongst other sources, in inputs pro-vided by social research. Neither Latin America nor the Ro de la Plata are exceptions to this rule.

However, emphasizing the need of improving the nexus between social sciences and publicpolicies in our countries does not imply the belief that technocracy should have supremacy overpolitics. It is not true that democracy and technocracy are necessarily opposite principles 1. Anactive and growing civic participation is not incompatible with the incorporation of expert knowl-edge in public policies. Actually, it is not difficult to grasp that democracy fragility grows whenelected Governments fail to solve, efficiently and effectively, the complex problems of economicand social development. Latin American political history during the 20th Century shows us that it isnot possible to build stable democracies with the social and economic restrictions caused byunderdevelopment and in the midst of the generalized poverty of the population.

In the third place, we would to anticipate ourselves to another criticism. Saying that thinktanks play an important role in modern societies does not necessarily imply questioning therole of political parties. In fact, it is true that think tanks often compete with parties, as Baierand Bakvis hold in their article. However, the most virtuous scenario is the cooperationbetween both of them. In fact, one of the main challenges for think tanks in Latin America is tohelp build strong political institutions and stable and competitive party systems. Latin America,since the times of the CEPAL (Comisin Econmica para Amrica Latina or EconomicCommission for Latin America), has invested a lot of energy thinking on social and economicproblems. It is essential to continue to make efforts to analyze and develop proposals that

might help solving the institutional, political and social problems of our region.

Lets go back to the content of the book, after these short but necessary comments. Thereader will easily notice that the articles of this book have a particular approach. Actually, wecould say that each one of them tackles the think tank topic from a different perspective. Aseditors of the book, we do not find this to be a problem. On the contrary, we trust this diversi-ty will allow the reader to have a very broad overview of several dimensions and different the-oretical and methodological approaches used by the authors in their researches.

We have organized the content of the book in three relatively homogeneous parts, since we

grouped the essays by common characteristics or approaches. The first part, Methodologicaland Conceptual Aspects on the Relationship between Think Tanks and Governments, gathersthree important articles. On the first place, we include a chapter of Donald Abelsons last book,A Capitol Idea2. Abelson is one of the most renowned specialists and the author of some of themain reference materials in this field. In this chapter, Abelson offers an excellent approach on theproblem of how to measure, with scientific rigor, combining quantitative data and a qualitative


1. The way democracy and technocracy complement each other has been carefully developped by Fernando Filgueirain La antiptica pero necesaria defensa poltica de la tecnocracia en Amrica Latina, in Vera, Miguel. Evaluacin parael desarrollo social: aportes para un debate abierto en Amrica Latina, Magna Terra Editores, Guatemala, 2006.2.Abelson, Donald,A Capitol Idea: Think Tanks And US Foreign Policy, McGill-Queen's University Press (September 2006).


8/319

perspective, the effective influence of think tanks in the public policy cycle. We decided to includethis article, precisely, because we consider that one of the most important challenges, both foracademics and for think tanks, lies on improving the mechanisms to measure their influence. LikeCarol H. Weiss and many other important specialists have stated, the influence of social researchis generally gradual, indirect. For that same reason, it is difficult to measure.

The second article is a relatively unknown work of Gerald Baier and Herman Bakvis, twoCanadian scholars. It analyzes the relationship between think tanks and political parties inmodern democracies. Including this article seemed essential to us because, no matter howimportant both institutions are in the present, there is very few written material about the rela-tionship between think tanks and political parties3.

Both chapters offer high quality methodological and conceptual tools that can be applied tothe study of think tanks, regardless of the specific political context.

The third chapter of this book belongs to Ajoy Datta, Nicola Jones and Enrique Mendizbal.It explores the common factors intervening in the development of think tanks influence onpublic policies, both in Latin America and other regions of the world. The article emphasizestwo aspects: the way think tanks manage to enter national political processes, and the formaland informal bonds think tanks establish with local and foreign key actors.

In the second part of the book, titled Think Tanks in Latin America, we decided to groupfour articles that give us a perspective of the reality of think tanks in Latin America. Like in thefirst part, the texts in this group do not overlap each other, but instead, they offer complemen-

tary perspectives. Miguel Braun, Mariana Chudnovsky, Nicols Ducot and Vanesa Weyrauchsarticle was prepared for the project Bridging Research and Policy from the GlobalDevelopment Network. The document was written after a comparative study of different poli-cy research institutes in Latin America, Asia and Eastern Europe. The authors analyze theendogenous and exogenous factors that affect think tanks influential capability. This docu-ment is valuable for more than one reason. On the one hand, the empirical aspect allows us tohave a comparative point of view of Latin American think tanks. On the other hand, in a theo-retical aspect, it provides many useful ideas on organizational solutions and concrete practicesto help maximize think tanks influence in public policies.

The second article, by Orazio Belletini, goes over the formation process of think tanks inLatin America and illustrates the influence of this kind of organizations in public policies by pro-viding recent examples of reforms in Latin America.

In the third place, Gerardo Ua, Carina Lupica and Luciano Strazza analyze think tanks par-ticipation in social policies, taking into account that participation in the public policy cycle in


3. There has been a recent effort to develop that line of investigation. We recommend the reading of a book edited by EnriqueMendizbal and Kristen Sample, called Dme a quin escuchas. Think tanks y partidos polticos en Amrica Latina, IDEA-ODI,Lima, 2009. Available online in: http://www.odi.org.uk/resources/details.asp?id=4204&title=think-tanks-partidos-politicos.


9/319

three of the most important conditional cash transfer programs: Progresa - Oportunidades(Mexico), Plan Jefes y Jefas de Hogar Desocupados (Argentina) and Chile Solidario (Chile).

The authors find that different levels and ways of think tanks participation might be affect-ed by the political and institutional context where the programs emerge, the governments

political will to create spaces for the interaction with think tanks and the proximity betweenknowledge centers and the public function.

Finally, lvarez-Rivadulla, Markoff and Montecinos study the attention given by two tradi-tional and influential American think tanks, Heritage Fundation and Cato Institute, to LatinAmerican reality. From a transnational point of view, this research emphasizes the bondbetween American organizations and certain Latin American think tanks, trying to understandhow these actors, who crate knowledge, work within the context of the relationship betweenthe North and the South of the continent. Unlike the first three articles, this last one is not theresult of the lessons of the daily management of think tanks, but the result of specific interrog-

atives that emerge from the academic world.

Our path takes us from general aspects to more specific ones. The third part of the book iscalled Think Tanks in the Ro de la Plata and it gathers study cases from Argentina andUruguay. Following a quantitative strategy (after Donald Abelsons methodological recommen-dations), Antonio Camou offers an exhaustive study on the relationship between think tanksand economic policies in Argentina, from 1985 to 2004, with a focus in the role of the experts.Gerardo Ua, on his part, introduces the determining factors that influence the development ofthink tanks in Argentina. His main argument is that, for these organizations, there is a tension

created between their increasing participation in the process of public policies and their ownpermanence in time.

The case of Uruguay is examined by the last two essays. In the first one, Adolfo Garc fol-lows Gerald Baier and Herman Bakvis concepts and describes the relationship between thinktanks and political parties in that country. He explains that, at least until the victory of the leftin 2004, there was no stable and broad interaction with university centers institutions thathave play the role of think tanks in Uruguay in the last 50 years. Despite lack of this kind ofinstitutions, political parties were not able to create important think tanks of their own.

The last chapter, by Javier Gallardo, Adolfo Garc and Paulo Ravecca, builds a broad mapof think tanks in Uruguay which enables the authors to describe the sources of funding, themagnitude of human resources and the outreach strategies that think tanks follow in order toinfluence public policies decision making. They also analyze the dynamics of the bondbetween the studied think tanks and the Frente Amplio Government between 2005 and 2008,seeking to identify circumstantial and structural determining aspects of that bond. By doing so,this article complements the former one.

* * *



10/319

This book is, in a certain way, a culmination. A culmination of a process that started afterthe incentive provided by UNESCOs MOST Forum. At the same time, we would like to thinkof it as a new starting point, of a path already initiated. It is evident that it is absolutely neces-sary to learn more about the relationship between social research and public policies, espe-

cially about think tanks. Many areas need a deeper look. What are the differences and similar-ities between think tanks in Latin American and in other parts of the world? Which are the maindifferences between the various national traditions inside Latin America? Which are the mostsuccessful examples of influence and which are the endogenous and exogenous factors thatmight explain those good performances? Which topics are prioritized by think tanks in theregion? Which issues have been less studied? In our region, what is the place, in the agendaof the main think tanks, of programs that fight poverty and inequality? What are the mainfundraising strategies? To what epistemic community do members of the technical teams ofthe think tanks of the region belong to? What are their professional profiles and their level ofacademic formation? What do political parties think of think tanks? Do they see them as allies

or as adversaries? On the other hand, what do think tanks think of political parties? What dothey do to strengthen political parties and institutions? Of course, this is not an exhaustive list.It is just an attempt to show the volume of the pending research ahead of us.

* * *

We would like to finish this Introduction expressing our sincere gratitude towards IDRC and

all the people that have cooperated with the publication of this book. To Federico Burone, fromIDRCs Regional Office, as well as Antonio Ciccioni, Andrea Puppo, Andrs Rius and MaraScherschener. The financial assistance of this institution made our publication possible. A veryspecial thank to Miguel Braun and Laura Zommer, from CIPPEC, who had an active participa-tion in this new version. To Patricia Draper, who translated several materials included in thisbook, and to Luciano Strazza and Dolores Arrieta, an essential support for the making and theedition of this project. And of course, our greatest gratitude to those who enriched with theirideas and experiences the edition of the book, and to those who wrote its different chapters.



11/319

PART I.

Methodological and Conceptual Aspects on the

Relationship between Think Tanks and Governments



12/319

1. Is Anybody Listening?

Assessing the Influence of Think Tanks

DONALD E. ABELSON1

The purpose of this chapter is to consider how scholars can better assess the influence or

impact of think tanks at different stages of the policymaking process. Although the behavior of

think tanks has been subjected to increased scrutiny in recent years, little progress has been

made in evaluating the nature and extent of their contribution to public policy. Indeed, rather

than discussing how different methodological approaches can be used to study think tank

influence, journalists and scholars have for the most part been content to make sweeping and

often unfounded observations about their policy impact. Needless to say, these have done lit-

tle to advance our knowledge of how think tanks engage with the public and with policymak-

ers to influence the political agenda.

To address what is clearly a significant shortcoming in the literature, this chapter will discuss

the advantages and disadvantages of relying on quantitative and qualitative approaches to

evaluating think tank performance. By so doing, we can begin to think more critically about

how to overcome some of the many methodological obstacles that limit our ability to make

informed observations about the influence of think tanks. Although the concept of influence is

ambiguous and difficult to grasp, it is central to any discussion about politics and policymak-

ing. It is also central to any discussion about think tanks and their efforts to become

entrenched in the policymaking process. As students in introductory political science coursesare reminded, politics is about the struggle for power and the ability of various individuals and

organizations to achieve desirable outcomes. But it is also about ideas, the ambition of lead-

ers, and the goals and aspirations of citizens. In short, it is about competing visions of the

national interest and the many forces that shape the fate of nations. It is for these reasons that

we need to explore more fully how think tanks contribute to policy development.

The American political system is the ideal environment for think tanks to inhabit. As organ-

izations competing in the free and open marketplace of ideas, they have innumerable oppor-

tunities to share and discuss their ideas with the public, with the media, and with policymak-

ers. However, as in any business, they must be able to monitor the quality and appeal of theirproducts. Unfortunately, unlike that of Fortune 500 companies and the millions of small busi-

nesses across the United States, the success of think tanks cannot be measured in terms of


1. Donald Abelson is Professor and Chair, Department of Political Science and Director, Centre for American Studies, atThe University of Western Ontario where he specializes in American Politics and U.S. Foreign Policy. Dr. Abelson is the

author of several books including, Do Think Tanks Matter? Assessing the Impact of Public Policy Institutes: Second

Edition (2009) and A Capitol Idea: Think Tanks and U.S. Foreign Policy (McGill-Queens University Press, 2006). He is

currently writing a study on the relationship between think tanks and US presidents. His work has also appeared in sev-

eral edited collections and academic journals, including: Global Society, Presidential Studies Quarterly, The Canadian

Journal of Political Science and Canadian Public Administration.


13/319

profits and losses. It can only be measured by assessing their ability to influence the views and

attitudes of the public and policymakers. Before delving into the various ways to measure influ-

ence, we must first discuss what scholars mean byinfluence. As we will discover, although the

process of wielding influence can be easily summarized using lines and arrows, understanding

the nature of influence and how it is exercised is far more complicated. Influence is often treat-

ed in a linear fashion whereby two parties negotiate policy outcomes but, in reality, a moreholistic approach may be required to comprehend how policy influence is achieved.

Easier Said than Done: the ABCs of Influence

In one of the more serious treatments of influence in the literature on international relations,

K.J. Holsti suggests that influence occurs when A convinces B to do X, convinces B not to do

X, or persuades B to continue a course of action or policy that is useful to, or in the interests

of, A. (Holsti, 1988: 142). As inability to achieve any of these desired outcomes, which could

be the result of a multitude of factors relating to either A or B, would suggest that under these

circumstances, A was unable to exercise influence. For Holsti, influence, which he regards as

an aspect of power, is essentially ameans to an end. Some governments or statesmen may

seek influence for its own sake, but for most it is instrumental, just like money. They use it pri-

marily for achieving or defending other goals, which may include prestige, territory, souls, raw

materials, security, or alliances (Holsti, 1988: 141).

The easiest way to measure influence, according to Holsti, is to study theresponses of those

in the influence relationship. If A can get B to do X, but C cannot get B to do the same thing, then

in that particular issue, A has more influence. If B does X despite the protestations of A, then wecan assume that A, in this circumstance, did not enjoy much influence (Holsti, 1988: 150). In other

words, influence is perceived to have taken place if B responds to or reacts in a manner accept-

able to A. Conversely, influence has not taken place, according to Holstis model, if B does not

abide by As wishes. In short, influence is tied directly to specific policy outcomes.


A BINFLUENCES

To do X

Not to do X

To maintain the status quo


14/319

In studying the relationship between states in the international system, as Holsti does, it

may make sense to employ a linear model of influence. After all, by understanding the power

capabilities of two states engaged in conflict and the efforts undertaken by either or both par-

ties to exercise influence, it might be possible to explain what factors may have helped or hin-

dered As and Bs efforts to achieve their goals. Indeed, if scholars are interested in evaluating

why some arms control negotiations between the United States and Russia succeeded andothers failed, this model could prove to be useful. However, relying on it to evaluate the influ-

ence or impact of think tanks and other non-governmental organizations in the policymaking

process is problematic.

To begin with, in studying the policymaking process and the various individuals and organ-

izations that participate in it, it is critically important to understand the identities of A and B. In

Holstis model, A and B are sovereign states that have considerable resources at their dispos-

al to exercise influence. Now imagine that A is a resident scholar at a Washington-based think

tank with expertise in foreign policy and security studies and B is the chair of the Senate

Foreign Relations Committee. A has published a study on national missile defense and is

asked to testify before this committee. According to Holstis model, if As recommendations are

not followed by B and in fact are rejected (in other words, if As goal is not achieved), A would

be perceived as having no influence. Alternatively, if As recommendations are embraced by B,

A would be deemed to have had influence. Yet, as we will discover, both scenarios raise seri-

ous methodological problems.

First, we cannot assume that if As recommendations are dismissed by B, A has had no influ-

ence in the policymaking process. Although A may not be able to take credit for influencing a spe-

cific policy decision, A could have played an important role in helping the public, policymakers,and the media to consider other approaches to resolving a potentially difficult policy problem.

Indeed, as Peschek and others have argued, think tanks are most effective at framing the param-

eters of public policy debates. Second, A may be well positioned to share ideas with policymak-

ers at different levels of government, who in turn could draw further attention to the issues A has

flagged. As a result of suggesting that influence is tied directly to policy outcomes, scholars are

ignoring the many access points that think tanks and other non-governmental organizations have

to the policymaking process. Conversely, by presuming that A has had influence over B because

As preferences are satisfied, Holsti allows himself to fall into a different trap. Even if the recom-

mendations A has proposed closely resemble policies that are introduced, he should not take for

granted that A has had influence over B. It is conceivable, as we will discuss, that A has onlyreinforced what was on Bs mind or that other domestic and external forces compelled B to

act in a certain way. In reality, despite the perception of having considerable influence over B,

A may have exercised very little.

Holsti should also keep in mind that, given the vast number of individuals and organizations that

compete in the marketplace of ideas, it is often difficult to identify the source and origin of an idea

which could span several generations. As scholars of public policy are well aware, every success-

ful idea has a hundred mothers and fathers. And since many ideas take years before they make

their way onto the political agenda, it is also likely that they have grandmothers and grandfathers.



15/319

Regardless of an ideas gestation period, one thing is certain: every bad policy idea is an

orphan.

Think tanks, as noted, have a vested interest in creating the impression that they exercise

enormous influence. If this were not a concern, they would not devote so much time and

resources to enhancing their profile. However, scholars must resist the temptation of drawingon anecdotal evidence to support or reject these claims. They must also resist the temptation

of trying to simplify a process that is anything but simple and straightforward. Influence, like

so many other aspects of politics, is far more complicated than the linear model suggests and

cannot be reduced to two or three likely outcomes. In fact, contrary to Holstis model, it is not

always possible to confirm when A has or has not had influence. As much as scholars would

like to argue with some degree of confidence that A has been able or unable to exercise influ-

ence over B, it is increasingly difficult to do so because of the complexity and ad hoc nature

of the policymaking process. Rather than concluding that individuals and organizations have

or do not have influence, scholars may want to consider the likelihood that participants in pol-

icymaking enjoy different degrees or levels of influence at different stages of the policy cycle.

In the following section, we will discuss how scholars can do so by embracing a more holis-

tic approach to the study of policy influence. In addition to offering an alternative to examining

influence in a linear fashion an approach that assumes that two players will rely on various

strategies to achieve their desired goals this model compels scholars to think of the policy-

making process as a series of conversations taking place (often simultaneously) between mul-

tiple actors in distinct policy environments. In this model, influence is not tied directly to spe-

cific policy outcomes but is achieved through the interaction and exchanges between various

participants who are directly and indirectly involved in the policymaking process.

A Holistic Approach to Policymaking

In reflecting on the influence of think tanks on the media, Congress and the Executive, Leslie

Gelb, the former president of the Council on Foreign Relations and Pulitzer Prize-winning cor-

respondent for the New York Times, commented that it is highly episodic, arbitrary, and diffi-

cult to predict.2 His remarks were limited to think tanks, but he could just as easily have been

making an observation about the nature of the policymaking process in the United States. It is

because policymaking is highly episodic, arbitrary, and difficult to predict that a holisticapproach to studying policy influence may prove more useful.

In some respects, this approach builds on the work of scholars who study policy or epis-

temic communities and issue networks in the United States (see, for instance: Heclo, 1978).

Policy communities and issue networks are composed of individuals and organizations that,

by virtue of their expertise in a particular policy area, are invited by policymakers to participate

in various stages of the policymaking process. These communities, which are divided into two


2. Interview with Leslie Gelb, 22 February 2005.


16/319

spheres the core (sub-government) and the periphery (the attentive public) seek to influence

specific public policies.3 The approach that we are considering is similar to policy communities

in so far as it focuses on different clusters or centers of knowledge and expertise. However,

unlike policy communities, which are created in large part to influence one or more government

policies (X), a holistic approach considers how multiple actors (represented by An, Bn, Cn, and

Dn) attempt to influence the environment in which policy decisions are made. In other words,a holistic approach to policymaking assumes that while officials in the White House, on Capitol

Hill, and in various government departments and agencies attend to the affairs of state, con-

versations are taking place between policy experts in universities, in think tanks, in interest

groups, and in the private sector which, with the assistance of the media and other outlets, can

help to enrich policy debates. While elected officials may prefer to insulate themselves from

the discussions taking place in or between clusters An, Bn, Cn, and Dn, they cannot ignore

how these conversations shape the political agenda

Among the many benefits of a holistic approach is that it compels scholars to think about

policy influence, not in terms of how it is exercised between two players, A and B, but how it

can be fostered over time by different individuals and organizations acting alone or working

together or in concert with various policymakers. Such an approach also provides a broader

and more sophisticated understanding of policy influence. Recall the example we used of a


An

Cn

Dn

Bn

X

3.A detailed discussion of policy communities is provided by Lindquist in Public Managers and Policy Communities.

(Lindquist, 1992: 12759).


17/319

resident scholar from a Washington-based think tank who was asked to testify on national mis-

sile defense to the Senate Foreign Relations Committee. Using Holstis linear model of influ-

ence, we suggested two possible outcomes, given As preference that the recommendations

put forward be adopted: A convinces the chair of the committee (B) to endorse the recommen-

dations, in which case As influence is achieved, or B rejects As recommendations, in which

case A exercises no influence. The linear model provides an all-or nothing proposition Aeither has or does not have influence.

A holistic model which acknowledges that influence can occur in different ways and at dif-

ferent stages of the policy cycle presents a more realistic pic-ture of how A might achieve influ-

ence. Even though A might not be able to convince B to endorse the proposed recommenda-

tions, As testimony may spark a debate in the media, in academic circles, in the Oval Office,

at other think tanks, and in countless other places where public policy is discussed and ana-

lyzed. The fact that A may not have altered Bs position regarding national missile defense does

not mean that A lacks influence; nor does it mean that in the medium or long term, As recom-

mendations will be ignored. Rather, Bs unwillingness to fulfill As wishes suggests simply that

in this instance, B is unable and/or unwilling to follow As advice.

The timeframe over which influence occurs is also an important consideration, particularly

with respect to matters of war and peace. Although Holsti does not specify what he considers

a reasonable time frame for A to influence B, it is clear from the linear model he presents that

once B makes a decision, A is no longer in a position to exercise further influence. Unable to

convince B to act according to As preferences, A may seek influence through other channels.

This was certainly the case when the Bush administration decided to deploy troops to Iraq

after it became clear that the United Nations Security Council would not endorse the invasion.Decisions such as the one leading to the overthrow of Saddam Hussein reinforce why it is

important to look beyond the narrowly defined parameters of a linear model of influence; a

holistic model encourages scholars to pay attention to what takes place after a decision is

made, a period in which residual influence may surface. Let me explain. If B does not act

according to As wishes, as the example mentioned above illustrates, it does not necessarily

mean that in the short, medium, or long term A will be denied influence. Moreover, if B does

act according to As wishes, it does not mean that A will immediately move on to the next issue.

Indeed, Bs accommodation of A may provide A with an incentive to influence other decision-

makers and stakeholders. In some respects, residual influence can be regarded as goodwill

which develops over time between individuals and organizations that have reaped tangiblebenefits through cooperation. This goodwill in turn could, as early functionalists such as David

Mitrany predicted, spill over into more sensitive areas of negotiation (Mitrany, 1996).4

Unlike a linear model, a holistic approach does not try to reduce influence to an all-or-noth-

ing proposition; as noted, it acknowledges that there are degrees and levels of influence. It also

acknowledges that there are individuals and organizations which, by virtue of their expertise and


4. On theories of functionalism and integration and how they can be used to explain the success of interstate coopera-

tion, see Mitrany,A Working Peace System.


18/319

connections to key policymakers, are well equipped and positioned to influence both the policy-

making environment and specific policy decisions. In a holistic approach, it matters if A is a schol-

ar at a think tank, the editor of a newspaper, the head of an interest group, a professor at one of

the nations leading universities, a CEO of a major corporation, or a former cabinet secretary.

In the final analysis, it is important to recognize that with any important policy issue, whether

it is strategic defense or the war on terror, there are literally hundreds of organizations, includ-ing think tanks, which try to convey their ideas to policymakers. And as scholars interested in

how policy agendas are shaped, it is our responsibility to determine the most effective ways to

evaluate their contribution to public policy.

A holistic approach to studying policy influence is not neat and tidy, but neither is the

process by which public policy is made and implemented. The efforts of think tanks to influ-

ence public policy cannot be explained by limiting our discussion to two players trying to nego-

tiate favorable outcomes; nor can it be explained through computer-generated flow charts. The

process by which think tanks and other non-governmental organizations try to assert influence,

like the process of making policy decisions, is, as Gelb reminds us, highly episodic, arbitrary,

and difficult to predict. Ironically, this is what makes the study of public policy interesting.

Understanding who exercises influence and under what conditions it is achieved is a con-

stant challenge for scholars in the field. Part of the challenge, in addition to developing an

appropriate model to study influence, is determining the most effective ways to measure poli-

cy influence. In the following section, we shift our attention to how scholars can use both quan-

titative and qualitative indicators to assess or evaluate the influence of think tanks. How use-

ful these criteria are for assessing the influence of think tanks on public opinion and public pol-

icy is a question that will undoubtedly lead to a lively exchange in academic circles. In the inter-ests of promoting such an exchange, we will begin by focusing on what think tanks seem to

value most making the headlines.

Can Policy Influence be Measured? A Quantitative Approach

The Media

On any given day, you can pick up a newspaper, turn on the radio, watch the news or a cur-

rent affairs program, or scan the thousands of political Web sites on the Internet to find outwhat is on the minds of policy experts at Americas leading think tanks. And if that is not

enough, you can subscribe to Think Tank Watch, a weekly e-mail newsletter prepared and dis-

tributed by the Canadian embassy in Washington which summarizes the research activities of

Washington-based think tanks. The Canadian high commission in London has recently begun

a similar newsletter to monitor the work of British think tanks.

For twenty-four hours a day, seven days a week, the print and electronic media in the United

States and around the world look to policy experts to shed light on important and controver-

sial political issues, and think tanks are only too willing to oblige. Often referred to as talking



19/319

heads, scholars from think tanks appear with great regularity on the network news and on

political talk shows to comment on a wide range of domestic and foreign policy issues. They

become particularly visible in the hours and days following historic events such as the terror-

ist attacks against the World Trade Center and the Pentagon, when journalists and media exec-

utives scramble to make sense of what has transpired. When they are not on television or on

radio, policy experts from think tanks share their insights on the Op-Ed pages of major Americannewspapers. In short, think tanks understand the demands placed on the media and respond by

providing them with a steady stream of timely and relevant information.

Think tanks, like corporations involved in the entertainment and advertising industries, appreci-

ate the impact the media have on shaping our perceptions of what and who is important. That is

why, as discussed in the previous chapter, they devote so much time and resources to strength-

ening their ties to various media outlets. Few think tank directors need to be reminded of the

potential benefits of generating substantial media exposure; positive exposure not only enhances

the credibility of think tanks, but even more importantly, it creates the impression that they wield

enormous influence, a valuable currency that can be used to achieve desirable outcomes.

Having influence is something all think tanks covet, but in reality, most settle for the percep-

tion of exercising influence. And what better way to create the perception of influence than to

capture the attention of the media. As the marketplace of ideas has become increasingly com-

petitive, several think tanks have closely monitored their media exposure relative to their rivals.

What is at stake is more than bragging rights; it is the opportunity to trans-late heightened

exposure into additional philanthropic, corporate, and private funding. In the following section,

we will examine the amount of media exposure a select group of think tanks with expertise in

foreign and defense policy generated in leading American newspapers and on the televisionnetworks between 2001 and 2005. The purpose of compiling these data is not to confirm

which think tanks do or do not have influence but to demon-strate how this indicator of influ-

ence may be used in quantitative evaluations of think tank performance.



20/319

Using the database Lexis/Nexis, we recorded the number of times a select group of think

tanks the Project for the New American Century, the Centre for Security Policy, the Hudson

Institute, the Carnegie Endowment for International Peace, the Cato Institute, the Heritage

Foundation, the Hoover Institution, the Brookings Institution, the Center for Strategic and

International Studies, the American Enterprise Institute, the Council on Foreign Relations, and

rand were mentioned in six major newspapers the Wall Street Journal, the Christian Science

Monitor, the Washington Post, the Washington Times, USA Today, and the New York Times in

relation to six important events or issues Iraq, Afghanistan, 9/11, al-Qaeda, terrorism, and

missile defense between January 2001 and January 2005 (see table 1 and tables A4.1 to

A4.6).5 The number of references to each think tank may be inflated since more than one ofthese terms could appear in a newspaper article, but our purpose here is to simply demon-

strate a general pattern in think tank exposure.


Think tank

PNAC

Hudson

Center forSecurity Policy

Hoover

Cato Institute

Carnegie EndowmentFoundation

AEI

Heritage

RAND

CSIS

Council onForeign Relations

Brookings

Total

Irak

60

104

147

324

258

398

633

544

391

742

853

1,054

Afganistn

28

50

31

125

121

187

181

245

260

327

358

495

11/9

37

97

81

211

272

204

401

386

528

362

469

786

Al-Qaeda

20

32

47

84

87

114

146

150

232

231

288

271

Terrorism

52

158

191

354

437

409

629

685

691

754

846

1,270

Anti missiledefense

12

30

66

38

31

115

61

119

63

73

71

169

Total

209

471

563

1,136

1,206

1,427

2,051

2,129

2,165

2,489

2,885

4,045

20,776

%

1.01

2.27

2.71

5.47

5.80

6.87

9.87

10.25

10.42

11.98

13.89

19.47

Note: Because of similarity of search criteria, there is likely some article overlap in the numbers.

Source: LexisNexis

SUBJECT

Table 1. Print Media Coverage of Selected Think Tanks, January 2001 - January 2005: Totals of all Media Sampled

5. These think tanks were selected because of their interest and expertise in for-eign and defense policy. I also selected

six major newspapers that cover American politics inside the Beltway very closely. The issues and events that were cho-

sen represent some of the most important challenges confronting American decision-makers in the last four years. In thearea of domestic politics, other critical issues, such as health care, energy, and education, could have been identified. As

noted, the purpose in compiling these data was to demonstrate a pattern in think tank exposure and to identify one indi-

cator that could be used by scholars interested in using quantitative approaches to study think tank influence.


21/319

The exact number of citations is not as important as how scholars might use these figuresto determine the extent of think tank influence. A think tank receives one media hit for each ref-

erence made to the events and issues selected.

Of the twelve think tanks sampled, the Brookings Institution ranked first, receiving over four

thousand citations, or 19.48% of all print media references, almost twice the share of expo-

sure generated by AEI, Heritage, and Rand. Other think tanks that attracted considerable

exposure included CSIS and the Council on Foreign Relations. The Project for the New

American Century, the think tank credited with influencing the Bush doctrine, ranked last,

attracting only 1% of media exposure. There was tremendous variation in the exposure think

tanks generated across newspapers and issue areas. For example, Brookingss exposureranged from a low of 10.26% in the more conservative Washington Times to a high of 26.17%

in the more liberal Washington Post. By contrast, the Heritage Foundation, known for its con-

servative leanings, attracted the most coverage, 21.57%, in the Washington Times and the

least coverage, 4.49%, in what many would regard as the more liberal New York Times.

Brookings was quoted most often with respect to Iraq and terrorism, whereas RAND, for

instance, was cited most for its views on issues relating to terrorism and 9/11.

The results were similar for the broadcast media (see table 2 and tables A4.7 to A4.10). ABC,

and CNN called on experts from the Brookings Institution or cited one of its studies far more


Think tank

Hudson

PNAC

Hoover

Center for Security Policy

Hoover

Cato Institute

Carnegie Endowment Foundation

AEI

Heritage

RAND

CSIS

Council on Foreign Relations

Brookings

Total

Irak

0

1

1

4

4

2

17

11

42

65

30

152

Afganistn

0

0

1

1

0

3

2

3

7

8

7

31

11/9

0

0

0

0

0

0

0

4

4

6

7

17

Al-Qaeda

0

0

0

1

1

1

1

1

10

7

33

23

Terrorism

0

0

1

4

4

6

6

10

25

36

51

68

Anti missiledefense

0

0

0

0

0

0

0

2

8

1

1

5

Total

0

1

3

10

9

12

26

31

96

123

129

296

736

%

0,00

0.14

0.41

1.36

1.22

1.63

3.53

4.21

13.04

16.71

17.53

40.22

Source:Vanderbilt Television News Archive.

SUBJECT

Table 2. Total Television Exposure of Selected Think Tanks, January 2001 January 2005: Totals of all Media Sampled


22/319

often than from any other think tank: it received 40.22% of all broadcast media citations in this

sample, more than twice the share recorded by its closest competitor, the Council on Foreign

Relations (16.71%), and almost ten times the exposure generated by the Heritage Foundation

(4.21%). The exception was CBS, which relied equally on RAND and Brookings (both received

25.44% of the networks coverage). Once again, PNAC attracted very little attention (0.14%),

a step up from the Hudson Institute, which was not the subject of any discussion by the fournetworks.

In our review of these figures, an obvious question to ask is, why do some think tanks attract

more exposure than others? While there are several factors, including the size of a think tanks

budget, size of staff, area of research, ideological orientation, and geographic location, which

could help to explain why some think tanks are cited more than others (Abelson, 2002: 90:106),

our concern is less with how and why think tanks attract media coverage and more with what

these figures tell us or do not tell us about the extent of their influence.

Scholars interested in using quantitative approaches to studying think tank influence tend

to focus on media coverage because it is relatively easy to measure. Although it is time-con-

suming, scholars can draw on different databases, including the one used in this chapter, to

compile information on how much exposure think tanks attract in the print and broadcast

media. In so doing, they can record how many times think tanks are quoted, the policy issues

they comment on most often, and the period of time over which they appear to enjoy the most

visibility. Scholars may also elect to undertake rigorous content analyses of newspaper cover-

age so that they can measure the number of column inches devoted to think tanks. Moreover,

should they be so inclined, they can keep track of whether the ideological leanings of think

tanks are identified liberal, conservative, Marxist, libertarian and if the institutes aredescribed in a positive or negative manner. All of this information can then make its way into

studies evaluating the impact of think tanks.

The major advantage of tracking media exposure is that it enables scholars to identify those

institutes that are most active or relevant in framing the parameters of important public policy

debates. For those interested in how think tanks become involved at the initial stages of the

policy cycle, when ideas from multiple sources are being articulated, assembling a list of

organizations making the headlines is critical. By following think tanks that are making the

news and the issues that they are addressing, scholars can begin to examine more closely

what additional steps they might take to promote their ideas. While some think tanks may becontent to have one of their scholars or studies mentioned in a newspaper, others will try to

use the media to garner public support for a new policy idea or initiative. As momentum for an

idea builds, think tanks can then rely on many of the channels described to capture the atten-

tion of policymakers. In short, for many think tanks, being in the spotlight is a necessary,

though not sufficient, condition for exercising influence.

Unfortunately, other than providing some initial information on which organizations are at the

forefront of policy debates, media exposure tells us very little about the nature and extent of

think tank influence. Indeed, contrary to what is said in the annual reports issued by various



23/319

think tanks and their directors to their boards of directors, media visibility should not be equat-

ed with policy influence. The fact that a think tank study or report is referred to in the

Washington Post, for example, does not mean that the public or policymakers will be swayed

by its contents. To suggest otherwise, we would have to be confident that the report was read

and understood by a sufficiently large cross-section of Americans. Even if it could be estab-

lished, through a public opinion survey, that a significant percentage of the American peoplewere aware of a specific think tank report mentioned in the Washington Post, it would be dif-

ficult to demonstrate, for the reasons stated earlier in the chapter, that there reportinfluenced

either public attitudes toward a particular issue or the views of policymakers. Moreover, even

if it could be confirmed that public opinion had shifted as a result of this report, we would still

have to contend with the many methodological obstacles limiting our ability to trace the ori-

gins of an idea to a policy outcome. It is for all these reasons that, as previously discussed, a

more holistic approach to studying policy influence could prove more useful. Rather than try-

ing to equate media visibility with policy influence, we should explore how policy issues dis-

cussed in the media, in the academic community, in the private sector, and in think tanks even-

tually make their way to policymakers.

It is also important to keep in mind that our data on media visibility ignore vital information

necessary to make informed observations about think tank influence. What the Lexis/Nexis

database search provides is raw numbers on think tank citations. What is left out is the con-

text in which comments by think tank scholars or references to think tank studies are made.

Furthermore, the figures included in our charts do not reveal the type and scope of coverage

think tanks receive. For instance, we do not know how many of the four thousand citations

Brookings recorded in the print media were on the front or back page of newspapers. For obvi-

ous reasons, this factor could have a profound impact on the number of potential readers whocome across references to this and other institutions. Moreover, we do not know which, if any,

articles about the work of think tanks generated the most interest. Using Nielsen ratings, tele-

vision networks can determine roughly how many viewers watched a particular program, a sys-

tem that may help scholars study the exposure of think tanks in the broadcast media. However,

it would be of little use to those looking at think tanks and the print media. Put simply, track-

ing the media visibility of think tanks may provide scholars with an important piece of the puz-

zle, but it is still only a piece. The entire puzzle can only be completed when more information

about the involvement of think tanks is known.

Congressional Testimony

In addition to monitoring how much media coverage they receive, think tanks pay close

attention to how often their scholars are invited to testify before congressional committees. A

list of scholars who have given testi-mony as well as the full text of their remarks is often avail-

able on think tank Web sites. The reason for this practice is obvious. Think tanks want to con-

vey the impression that they are credible and important actors in the policymaking communi-

ty, and what better way to do this then to advertise their accomplishments? Although there are

several factors, according to Andrew Rich and Kent Weaver, that could account for why some



24/319

think tanks appear before legislative committees more than others6, we need to consider how

data on congressional testimony can be used to further our understanding of the involvement

and impact of think tanks in the policymaking process.

Between January 1st 2001 and January 1st 2005, the same time frame used to track media

exposure, policy experts from the think tanks sampled in our study testified 120 times beforeseven Senate and House committees with responsibility in the areas of foreign and defense

policy (see Table 3 and figures A4.1 toA4.6).7 Ranking first in number of appearances made was

the Center for Strategic and International Studies, a veritable whos who in the foreign policy-

making establishment. Experts from CSIS appeared before legislative committees on 33 sep-

arate occasions, or 27.5% of the time, well ahead of its closest competitors, the Brookings

Institution (22), AEI (14), and RAND (10). Only a handful of appearances were made by experts

at Cato (1), the Hoover Institution (2), PNAC (3), the Hudson Institute (4), the Center for Security

Policy (5), and Heritage (8).

According to our data, CSIS was most visible in the Senate, where its scholars appeared 20

times. Most of the testimonies were made before the Senate Foreign Relations Committee

where CSIS experts discussed issues ranging from bio-terrorism and the threat of infectious

diseases to the crisis in Iraq and negotiations with North Korea. In the House of

Representatives, CSIS maintained a strong presence (13 appearances), but fell slightly behind

the first-ranked Brookings Institution, whose scholars testified 14 times, mostly before the

House Committee on International Relations. AEI and Heritage put in strong showings with 10

and 8 appearances respectively.


6. Rich and Weaver argue that think tanks located inside the Beltway and those that are ideologically compatible with the

majority party in Congress are more likely to be called upon to testify before legislative committees. See their study Think

Tanks, the Media and the Policy Process, and Rich, Think Tanks as Sources of Expertise for Congress.

7. The study looked at the total number of appearances by scholars from twelve think tanks before the following com-

mittees: Senate Foreign Relations Committee, Senate Intelligence Committee, Senate Armed Services Committee, House

Permanent Select Committee on Intelligence, House Committee on Armed Services, House Committee on InternationalRelations, and House Committee on Homeland Security.


25/319

The data on congressional testimony raise similar methodological problems to that com-piled on media exposure. What we are presented with is information that may help us to iden-

tify think tanks deemed to be credible (at least in the eyes of some policymakers) but scarce-

ly any insight into which presentation or presentations given by think tank scholars were con-

sidered useful and relevant. Once again, the figures revealed in the charts do not speak to the

issue of influence but merely address the frequency with which think tanks are called upon to

testify before Congress. And as Rich and Weaver point out, there are several factors that could

explain why some think tanks appear more regularly before legislative committees than others.

Without paying close attention to the committee proceedings in which scholars from think

tanks testified, it is virtually impossible to predict how much or little influence they exercised.Indeed, in the absence of detailed information about the inner workings of individual commit-

tees and the policy preferences and goals of its members, one could reasonably conclude that

data on think tank testimony have limited utility. However, if one compares such data to that

on media exposure, it becomes clear that this approach to assessing policy influence may

prove more promising. In fact, by comparing the visibility of think tanks before Congress and

in the media, we can observe that they enjoy different levels of recognition at different stages

of the policy cycle. For example, while the Brookings Institution received 40.22% and 19.47%

coverage respectively in our sample of the broadcast and print media, its scholars appeared

18.3 % of the time before Congress. In contrast, while CSIS enjoyed the strongest presence


Think tank

Cato Institute

Hoover

PNAC

Hudson


Heritage



RAND

AEI

Brookings

CSIS

Total

Number of testimonies

1

2

3

4

5

8

9

9

10

14

22

33

120

Percentage

0.83

1.67

2.50

3.33

4.17

6.67

7.50

7.50

8.33

11.67

18.33

27.50

Source: Source:LexisNexis.

Table 3. Combined Testimony before US Senate and House of Representatives Committees by Selected Think Tanks,

January 2001 - January 2005


26/319

before Congress (27.50%), its exposure in the broadcast (1.63%) and print (11.98%) media

was far less significant. By keeping track of where in the policy cycle some think tanks appear

to be most active, we can then explore in more detail the nature of their contributions to spe-

cific policy debates. In short, in assessing policy influence, what is important to highlight is not

which think tanks ranked first, second, or last in such categories as media exposure and con-

gressional testimony, but which think tanks, relative to other institutes and organizationsinvolved at the same stage of the policy cycle, were best equipped and positioned to influence

public policy.

We cannot determine this by focusing solely on quantitative indicators of policy influence.

Media exposure, congressional testimony, and other measurements such as number of publi-

cations produced and size of staff and budgetary resources can only take scholars so far.

Although data on these and other aspects of think tank activity can be useful in documenting

patterns and trends in the institutes behavior, a more comprehensive understanding of their

influence is required to probe more deeply into their involvement in the policymaking process.

The Invisible Cloak: Think Tanks, Public Policy, and Qualitative Indicators of Influence

Studying public policy and the efforts of non-governmental organizations to shape it would

be so much simpler if political scientists could magically transform their academic robes into

invisible cloaks. By becoming invisible, they could make their way around the White House,

Capitol Hill, and doens of government departments and agencies as easily as Harry Potter did

in the cavernous hallways and staircases of Hogwarts, the fictional school where he and his

friends were sent to learn the arts of witchcraft and wizardry. And like the bespectacled wizardwhose presence went undetected as long as he remained under his invisible blanket, political

scientists could travel effortlessly inside the corridors of power. With their cloaks around them,

there is little scholars could not ascertain about who and what was influencing the behavior of

policymakers. They could observe meetings between high-level officials in the Oval Office,

overhear phone conversations between key members of Congress, and gain access to confi-

dential documents. The mysterious and complex world of policymaking would suddenly

become open and transparent.

But the world at Hogwarts is very different from the one that policymakers in Washington

inhabit. At Hogwarts it is outsiders such as Harry Potter who can draw on extraordinary pow-ers to uncover what is taking place behind closed doors. By contrast, in Washington, insiders

use their authority as elected and appointed leaders to conceal the inner workings of the pol-

icymaking process. It is they, not the scholars who study them, who hide behind their invisible

cloaks. There are wizards in Washington the Washington Wizards but they play in the

National Basketball Association; they are not members of ancient societies hoping to uncover

the States most highly guarded secrets.

Since scholars cannot hide behind imaginary cloaks to observe first-hand why certain pol-

icy decisions were made, they must find other ways to shed light on the policymaking process.



27/319

We have already considered some quantitative indicators that may be employed to assess the

contribution of think tanks to policymaking. Among other things, we discovered that numbers

reflecting the amount of media exposure think tanks generate and/or the frequency with which

their experts appear before legislative committees often leave us with more questions than

answers. To remedy this problem, we may want to consider how qualitative approaches can

be used to provide further insight into how think tanks become involved in policymaking andthe nature and extent of their influence.

When scholars refer to qualitative approaches to the study of policymaking, what exactly do

they mean? Generally, they are referring to non-statistical methods of analysis, including

archival research and interviews, which enable them to reveal the inner workings of the policy-

making process. Those employing this approach maintain that while quantitative indicators

may be useful in evaluating policy influence, so too are intangibles such as personal contacts

with high-level officials who may have a profound impact on shaping policy decisions. For

example, in looking at how often experts from think tanks testified before congressional com-

mittees, we focused primarily on which institutes logged the most and fewest number of

appearances. No consideration was given to the reputation and standing of the experts who

testified or to the composition of the congressional committees. Why is this aspect important?

It is important because in the policymaking process, it matters who is providing advice and to

whom the advice is directed. It makes a difference if testimony is being presented by a former

cabinet secretary or a relatively unknown academic from a Washington-based think tank who

was called upon at the last minute to fill in for a more seasoned scholar. It also makes a differ-

ence whether policy experts are testifying before high-profile or relatively obscure committees,

and it makes a difference whether the topic being discussed is a priority for Congress and the

Administration or an issue that, like so many others, is forgotten overnight.

As we will discuss in some detail in our case studies, the relationships and contacts that

develop between think tanks and policymakers can often explain why some think tanks are

able to enjoy considerable access to various stages of the policymaking process. Heritage

president Edwin Feulners friendship with key members of the Reagan transition team in 1981

certainly played an important role in allowing the foundation to make its Mandate for

Leadership study known to the incoming administration. The same can be said for PNAC,

which relied on several of its more prominent members, including Secretary of Defense Donald

Rumsfeld and Vice-President Dick Cheney, to communicate the recommendations outlined in

its much publicized report Rebuilding Americas Defenses to President Bush.

However, while personal contacts and connections to decision-makers may help facilitate

access to various levels of government, they do not guarantee that policy experts from think

tanks and other organizations will be able to exercise policy influence. As Zbigniew Brzezinski,

national security adviser to President Carter, recently acknowledged, different communities of

scholars and former policymakers in Washington attempt to influence public policy. There are

some communities that do not have much influence, except perhaps in shaping public atti-

tudes to some extent through Op-Eds, television and so forth (...) and there are other commu-

nities [of policy experts who] have acquired a certain degree of public recognition [and] have



28/319

some degree of influence, not excessive, but some (...). [They have] influence because of their

natural access to policymakers.8According to Brzezinski, experts who fall into the latter cat-

egory can enjoy a great deal of influence when government policy gets into deep trouble. Take

Vietnam, for example, if things are not going well, all of a sudden viewpoints which are diver-

gent from the standard policy gain a great deal of circulation and are listened to much more

carefully.9

Although there may be an incentive for policymakers to listen to more diverse pointsof view when their policy gets into deep trouble, there are several other factors that may

explain why some administrations would be more inclined to turn to outside policy experts, a

subject that we will explore in the next two chapters.

The importance of qualitative analyses to the study of think tanks cannot be overempha-

sized. If done properly, archival research and interviews can produce volumes of information

that can document in great detail the critical factors which shape public policy. But quantita-

tive methods should not be overlooked or ignored. On the contrary, qualitative approaches can

offer scholars something that raw data cannot the historical and political context in which pol-

icy decisions were made. In chapters 8 and 9 we will draw on both qualitative and quantitative

approaches in order to evaluate the extent to which a select group of think tanks were able or

unable to exercise influence at different stages of the policy cycle. In so doing, we will remain

cognizant of the many methodological barriers often encountered in studying policy influence.

Assessing policy influence is inherently difficult, but it is necessary if we are to make any

progress in studying the involvement of think tanks and other non-governmental organizations

in the policymaking process. The alternative to rely on anecdotal information or data that tell

us little about whether policymakers and the public are listening to the steady stream of infor-

mation being distributed by think tanks is not an option.


8. Interview with Zbigniew Brzezinski, 20 May 2004.

9. Ibid.


29/319

Appendix


Think tank


Hudson

PNAC

Cato Institute

Hoover

Heritage


RAND

AEI

CSIS


Brookings

Total

Irak

3

8

17

24

42

53

60

104

119

124

246

247

Afganistn

0

1

6

12

11

17

22

64

34

60

102

106

11/9

0

7

9

41

28

35

39

137

76

58

140

164

Al-Qaeda

0

2

4

10

11

6

17

45

29

29

72

66

Terrorism

1

10

10

44

41

60

68

179

120

136

262

291

Anti missile

defense

4

2

4

5

6

4

26

26

8

18

24

43

Total

8

30

50

136

139

175

232

555

386

425

846

917

3,899

%

0.21

0.77

1.28

3.49

3.57

4.49

5.95

14.23

9.90

10.90

21.70

23.52


Source: LexisNexis.

SUBJECT

Table A4.2 Print Media Coverage of Selected Think Tanks: New York Times

Think tank

PNAC


Hudson

Hoover


RAND

Cato Institute

AEI

Heritage

CSIS


Brookings

Total

Irak

1

5

9

16

37

25

34

71

48

104

104

138

Afganistn

2

1

6

11

17

20

27

20

29

46

49

75

11/9

2

1

7

14

22

37

29

44

45

39

50

106

Al-Qaeda

0

0

3

5

13

15

16

14

20

26

37

28

Terrorism

3

3

10

28

34

57

50

63

72

89

93

161

Anti missiledefense

1

2

4

2

12

1

6

10

16

10

5

18

Total

9

12

39

76

135

155

162

222

230

314

338

526

2,218

%

0.41

0.54

1.76

3.43

6.09

6.99

7.30

10.01

10.37

14.16

15.24

23.72


Source: LexisNexis.

SUBJECT

Table A4.3 Print Media Coverage of Selected Think Tanks: USA Today


30/319


Think tank

PNAC


Hudson


Cato Institute

Heritage

Hoover

Brookings

CSIS

AEI


RAND

Total

Irak

2

5

12

12

25

25

38

51

60

62

64

41

Afganistn

0

0

4

12

6

12

18

15

25

18

16

23

11/9

1

2

9

10

21

22

25

37

33

43

42

98

Al-Qaeda

1

1

2

7

3

6

13

12

12

18

25

13

Terrorism

1

6

14

18

26

34

32

40

59

60

63

73

Anti missiledefense

0

2

0

10

0

1

1

9

4

7

2

3

Total

5

16

41

69

81

100

127

164

193

208

212

251

1,467

%

0.34

1.09

2.79

4.70

5.52

6.82

8.66

11.18

13.16

14.18

14.45

17.11


Source: LexisNexis.

SUBJECT

Think tank


Hudson

PNAC

Hoover

Cato Institute

Heritage


AEIRAND

CSIS


Brookings

Total

Irak

6

21

26

33

43

66

167

190107

188

213

342

Afganistn

3

10

11

10

21

28

77

4475

90

92

188

11/9

8

19

14

21

50

60

80

111132

104

124

298

Al-Qaeda

3

4

11

8

14

10

38

4068

79

76

94

Terrorism

10

30

25

33

81

96

157

161201

206

214

445

Anti missiledefense

3

6

6

5

6

20

39

1715

21

23

40

Total

33

90

93

110

215

280

558

563598

688

742

1,407

5,377

%

0.61

1.67

1.73

2.05

4.00

5.21

10.38

10.4711.12

12.80

13.80

26.17


Source: LexisNexis.

SUBJECT

Table A4.5 Print Media Coverage of Selected Think Tanks: Washington Post

Table A4.4 Print Media Coverage of Selected Think Tanks: Wall Street Journal


31/319


Think tank

PNAC


Hudson

RAND


Cato Institute


AEI

CSIS

Brookings

Hoover

Heritage

Total

Irak

9

47

42

55

118

90

122

142

164

133

179

295

Afganistn

7

27

26

28

49

35

25

54

69

64

72

136

11/9

8

23

36

49

56

99

69

97

77

102

114

186

Al-Qaeda

3

13

16

23

37

28

42

33

48

38

44

92

Terrorism

10

62

76

85

122

188

167

183

177

188

206

400

Anti missiledefense

1

12

12

12

14

12

53

11

15

34

20

66

Total

38

184

208

252

396

452

478

520

550

559

635

1,175

5,447

%

0.70

3.38

3.82

4.63

7.27

8.30

8.78

9.55

10.10

10.26

11.66

21.57


Source: LexisNexis.

SUBJECT

Think tank

CSIS

Hudson

Hoover

Cato Institute

PNAC

Heritage


AEI


RAND


Brookings

Total

Irak

0

0

0

1

1

1

2

9

17

15

33

57

Afganistn

0

0

0

0

0

0

0

1

0

2

4

15

11/9

0

0

0

0

0

0

0

0

0

1

1

7

Al-Qaeda

0

0

0

0

0

0

0

0

1

4

3

6

Terrorism

0

0

0

0

0

0

0

3

5

13

15

21

Anti missiledefense

0

0

0

0

0

1

0

0

2

0

1

3

Total

0

0

0

1

1

2

2

13

25

35

57

109

245

%

0.00

Documents

Garce, A., Uña G. (2010) Think Tanks and Public Policies in Latin America,