(SSH) Research and Science, Technology and

Comisión Sectorial de Investigación Científica (CSIC) Universidad de la República (Uruguay)

Social Sciences and Humanities (SSH) Research and Science, Technology and Innovation (STI) Policy-Making in Latin America: a Nexus Perception Study

Michele Snoeck and Judith Sutz with Claudia Cohanoff and Natalia Grass CSIC – Universidad de la República (Uruguay)

December 2010



Isabel Bortagaray collaborated in the first phase of the project.



Table of contents Introduction ................................................................................................................................ 2 1.

Conceptual approach to the research-policy nexus ................................................................ 3 1.1 SSH research and policy-making: shall the twain ever meet? ............................................... 3 1.1.1 Basic roles attributed to SSH research in public policy-making................................. 4 1.1.2 Some recent typologies of research to policy frameworks ....................................... 7 1.1.2 Studies focusing on developing countries ............................................................... 15 1.2 The role of SSH research in innovation policy development ............................................... 18

2.

Empirical evidence on the research-policy nexus in STI ........................................................ 21 2.1 Brief characterization of SSH innovation research in LA countries...................................... 22 2.2 Researchers’ perception of the use of their research in the PMP ....................................... 25 2.2.1 How do researchers pretend and intend influencing policy-making? ..................... 25 2.2.2 Can researchers pinpoint policy impacts of specific studies? ................................. 28 2.3 Researchers’ view of the obstacles to a closer research-policy nexus ................................ 30 2.3.1 Limitations on the research side ............................................................................. 32 2.3.2 Mismatches between research supply and demand ............................................... 34 2.3.3 The nature of the policy-making process ................................................................ 36 2.3.4 Governability and governance issues ...................................................................... 36 2.3.5 Influence of Northern frameworks on STI policy and research ............................... 38 2.4 Policy-makers’ view and use of SSH research on innovation .............................................. 39 2.4.1 Role assigned by PM to SSH research in policy-making .......................................... 39 2.4.2 Knowledge inputs used in STI policy design: PM’ versus researchers’ view ............ 41 2.5 Strong and weak nexus: an illustration ............................................................................... 47 2.5.1 Local productive and innovative systems in Brazil: an example of good match between researchers and policy makers ................................................................. 47 2.5.2 SSH research and STI policy-making in Venezuela .................................................. 49

3.

Conclusions and policy recommendations ........................................................................... 53 3.1 Overall conclusions from the empirical work ...................................................................... 53 3.2 The importance of research policy demand for evidence-informed STI policies................. 57

Bibliographic references ..................................................................................................................... 66 Annex 1 – Acronyms and abbreviations .............................................................................................. 69 Annex 2 – List of interviewed researchers and policymakers ............................................................. 70 Annex 3 – Questionnaires used for the survey ................................................................................... 75

2

Introduction The task of the EULAKS project to which this study responds consists in analysing the role and status of Social Sciences and Humanities (SSH) in designing and implementing innovation policies. Implicit in this task is the assumption that there are both encounters and mismatches between SSH research and policy development in the field of innovation, and that a better understanding of the interface between these two processes would help to conceive ways to foster the nexus. The objective of this study was thus to explore the science, technology and innovation (STI) research-policy nexus, based on the perception of this nexus by researchers and policymakers (PM). Capturing the complexities of this interface is essential to answer such questions as: How could PM make a better use of all that SSH research has to offer, from an innovation perspective? How could SSH researchers be more engaged in the development of public policies for innovation? What are the challenges of bringing both sides together with a view to matching innovation policies and the countries’ development needs?

The study involved a literature review on the SSH research-policy linkage, which is presented in chapter 1. Chapter 2 focuses on the results of a survey consisting in the conduction of 50 interviews to researchers and PM from a series of Latin American countries (Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Mexico, Uruguay and Venezuela) and to a few selected researchers from developed countries (Denmark, Netherlands, Spain, Sweden, United Kingdom, and United States). In chapter 3 we put forward our conclusions as well as tentative suggestions to improve the SSH research – STI policy nexus.

3

1.

Conceptual approach to the research-policy nexus

Research contributes to knowledge and the link between knowledge and policy has been looked at from several perspectives and disciplines (Box 1.1). The specific role of SSH research in policy design and implementation has been studied from a broad policy perspective and in some particular sectors, especially health and education. With time, the intricacy of the research-policy nexus has given rise to increasingly sophisticated and holistic models to account for the dynamics involved, mainly in the developed world. In contrast, from an innovation policy perspective, the use of SSH research in policy-making does not seem to have been dealt with in the same integral way, even less so in developing countries. In the present chapter, we briefly retrieve from the international literature those concepts, factors, and models, that should or could be taken into account when analysing empirical work on the SSH research and innovation policy interfaces. Box 1.1 – The knowledge-policy link from different perspectives "There is a great variety of areas of study that look into different aspects of the link between knowledge and policy, and many of them have, to a greater or lesser extent, been applied in developing country contexts. For example, ‘bridging research and policy’ [ODI-RAPID project] focuses largely on how to increase the uptake of research in policy-making, based on the conception of a ‘gap’ between the two based on cultural differences between researcher and policymaker communities, and resulting in a breakdown of communication, supply and demand of knowledge, etc. (Court, Hovland and Young 2005, Livny et al 2006). Innovation systems models focus on how existing knowledge ‘flows’ around a system (usually looking at the national level), looking at how factors such as infrastructure, networks, intermediary organisations and knowledge users and producers interact to affect how ideas and knowledge diffuse and are taken up in practice (e.g. Rath and Barnett 2006, Jones et al 2009). Political science is relevant for understanding how political behaviour affects the process of making policy, and in turn the role that knowledge has in this behaviour. Knowledge management looks at the processes and practices of the creation, acquisition, capture, sharing and use of knowledge, skills and expertise (Ferguson, Mchombu and Cummings 2008), which represents another perspective on the links between knowledge and policy. There is very little work that has attempted to bring together these perspectives at a theoretical level … , let alone with a focus on developing countries" (Source: Jones, 2009: 10).

1.1

SSH research and policy-making: shall the twain ever meet?

Policy design and implementation is indisputably a complex and dynamic process. Policies are only rarely the outcome of a rational and systematic approach to problem solving; they are not selected "by benevolent planners (or similar constructs), but are instead the outcome of strategic interactions among a number of key participants (voters, economic interest groups, politicians, technocrats) each with its own motivations and incentives… *These actors+ interact in a variety of formal and informal arenas (such as Congress or the street), which can in turn be more or less transparent" (Ardanaz, Scartascini, Tommasi, 2010: 2, 8). The multifactorial and multiactorial attributes of the policy-making process (PMP) thus prevent from considering the uptake of research findings in a linear and rational way, as if policymakers were waiting for research inputs to fill identified knowledge gaps. On the other hand, research is a complex world on its own that rarely delivers ready for use products. However, it often influences policy in intangible, indirect, lengthy, and sometimes indiscernible ways. How does interaction then develop between the research and the policy spheres? Based on a literature review, we will first define the meaning of research use in terms of policy development and then turn to some of the models that try to capture and picture the nexus.

4 1.1.1

Basic roles attributed to SSH research in public policy-making

Following the pioneer work of Weiss (1979) on the many meanings of research utilization 1, and the further elaboration on this issue in the international literature, most scholars now seem to agree in distinguishing three basic types of uses of SSH research in policy development: Instrumental use: research findings are directly applied to the solution of policy problems; they are a clear and identifiable input in the PMP. Though this is not considered in the literature to be the most common impact of research, the current evidence-based policy (EBP) trend tends to focus on the instrumental use of research (Box 1.2).

Box 1.2 – Research and evidence-based policy-making In the last two decades, there has been a growing interest in, and practice of, evidence-based policy (EBP) in various fields and countries (especially UK). Davies (2004:3) defines EBP as an approach that “helps people make well informed decisions about policies, programmes and projects by putting the best available evidence from research at the heart of policy development and implementation. This approach stands in contrast to opinion-based policy, which relies heavily on either the selective use of evidence (e.g. on single studies irrespective of quality) or on the untested views of individuals or groups, often inspired by ideological standpoints, prejudices, or speculative conjecture". As Marston and Matts (2003: 144) note: "It is difficult to imagine anyone arguing that policy should be based on anything but the best available evidence. The concept of EBP has an intuitive, common sense logic … *it+ acts as a catch phrase for 'scientific', 'scholarly', and 'rationality' …". However, the meaning of EBP has been contested. On the one hand, "[t]here is a risk that EBP will become a means for policy elites increase their strategic control over what constitutes knowledge about social problems in a way that devaluates tacit forms of knowledge, practice based wisdom, professional judgment, and the voices of ordinary citizens" (Ibid: 158). On the other hand, "[t]ensions in EBP discourse concern the relative value of research and other kinds of evidence as inputs into policy-making. This debate can be characterized as a continuum, with the rational actor model of policy, where research plays a major role in policy development at one end; and the political model, where research is just one input in the policy process –and often no the most influential– at the other (Cook, 2001). Of course, many social scientists and policy-makers sit somewhere near the middle of this continuum. Arguing that policymaking is inherently political, Nutley et al (2002:2, 145) prefer the term 'evidence-influenced' or 'evidence-aware' as a more realistic view of what can be achieved". Nutley et al. (2010: 133) situate the evidence and policy issue in the broad perspective of knowledge societies: "… the debate did take on a more ‘utilitarian turn’ in the 1990s and became more focused on the role of research in identifying what works (Solesbury, 2001). The factors that have contributed to this ‘turn’ include societal and economic shifts towards a more systematic creation and use of knowledge, easier access to previously scattered knowledge bases through technological developments and benchmarking, and the wider impact of the New Public Management (NPM) school of thought with its drive to reform and modernise government and public services in order to provide value for money (Duke, 2001; Mulgan, 2003). EBPP developments can be viewed as part of this developing broader discourse on the role of knowledge in societies and the implications this has for organisations within these societies. Not only have there been calls to develop more knowledge-intensive firms, but also demands for knowledge-intensive states and knowledge-intensive public services (Delvaux and Mangez, 2008)" (Ibid: 133). This utilitarian turn included renewed expectations about the usefulness of social sciences for policy-makers, as was expressed by a British Education Secretary in 2000: "we need to be able to rely on social science and social scientists to tell us what works and why, and what types of policy initiatives are likely to be most effective" (Delvaux and Manguez, 2008: 105). A similar concern is noticeable in the United States, as attests a recent report of the National Science and Technology Council of the Executive Office of the President (NSTC, 2008). It is claimed that: “Investments in many strategically important frontiers of science and engineering and the allocation of Federal resources across a complex and decentralized national research and development (R&D) portfolio must be guided by the best data and analysis available. A ‘science of science policy’, first named by the President’s Science Advisor, Dr. John Marburger, must be developed” (Ibid: 3).

Conceptual use or 'enlightenment' function: concepts and theoretical perspectives engendered by SSH research permeate or 'percolate' the policy-making process, and contribute to modify 'frameworks of thought' or 'the terms of the debate' on specific issues. 1

Weiss presented the following seven meanings or 'models' associated with the concept of SSH research use: knowledge-driven, problem-solving, interactive, political, tactical, enlightenment, and research as part of the intellectual enterprise of the society (Weiss, 1979).

5 In other words, the accumulation of knowledge through research causes changes in the perceptions and understanding of PM, thereby modifying the parameters and paradigms within which policy solutions are sought. A recent MOST-UNESCO policy paper (Milani, 2009) states: "It is useful to note that research for policy is not so much about providing answers as about changing the way questions are understood, so that people (researchers and policy makers, but other publics too) can begin to think differently, thus critically building the contours and contents of social problems”. Symbolic, strategic or legitimative use: research is used to support a predetermined position of decision-makers, to legitimize existing views, or, more generally, as 'ammunition' for political purposes. Also included in this category is a tactical use for purposes not related to the substance of the research but rather to the mere conducting of research (e.g., gaining prestige or allowing delays in decisions). Both the instrumental and conceptual uses of research actually underlie a common view hold by SSH researchers on their contribution to the PMP: 'helping government think' (Weiss, 1992; Estebanez, 2007) through a number of means, such as: Improving the level of information –in quantity and quality– in governmental entities. Performing studies and consultancy work commissioned by the government. Contributing with concepts and methodologies for analysis and evaluations; in particular, increasing the use of cognitive maps 2 on public issues. Participating in public debates, commissions, working teams or networks, etc., with informed opinions, thereby enriching the perspectives taken into account and, eventually, facilitating agreed-upon views on preferred strategies. Some authors consider that the concept of ‘research influence’ −i.e. the ability or power to produce effects in given areas (Kirkhart, 2000)− gives a better account than ‘research use’ of the ability to produce effects of a varying intensity. Accordingly, Knott and Wildavsky (1980) suggested the following 'stages' in research use: Reception: research reaches a policymaker's desk. Cognition: research is both read and understood. Discussion: research is discussed in meetings. Reference: research is cited in reports/presentations. Effort: some effort is made to favour the use of research. Adoption: research has had a direct influence on decision-taking. Implementation: research has been translated on the ground. Impact: the policy is successful: there are tangible, identifiable benefits in society. Unsurprisingly, empirical evidence has been easier to find for the first stages (conceptual uses) than for the last ones (instrumental use) (Nutley et al, 2007: 49). This confirms the importance of the conceptual meaning of research utilisation, while it also suggests difficulties associated with the instrumental use of research, as will be later developed. Lindquist (2001) argues that assessing research influence is typically about discerning intermediate influence, such as "expanding capacities of chosen actors and broadening horizons of others that comprise a policy network", and he distinguishes the following types of influence on policy (Ibid: 2324): Expanding policy capacities, by improving the knowledge/data of certain actors; supporting recipients to develop innovative ideas; improving capabilities to communicate ideas; or developing new talent for research and analysis. Broadening policy horizons, by providing opportunities for networking/learning within the 2

Cognitive mapping refers to those processes that facilitate the adquisition, codification, storing, remembering and manipulation of information by people on the nature of their environment.

6 jurisdiction or with colleagues elsewhere; introducing new concepts to frame debates, putting ideas on the agenda, or stimulating public debate; educating researchers and others who take up new positions with broader understanding of issues. Regarding the latter, the movement of people from one place to another is, indeed, an important factor: "A researcher becomes the head of a government bureau and brings his research knowledge with him. The head of a government agency serves on the advisory committee to a research study and absorbs the findings in situ, putting them to use back at the agency when a relevant situation arises" (Weiss, in Carden, 2009: ix). Affecting policy regimes, by modification of existing programs or policies; or fundamental redesign of programs or policies. These three categories were reaffirmed in IDRC’s evaluation of development research projects around the world, in very different political contexts (Carden, 2009). It was noted that "the most meaningful and lasting influence is less about specific policy change than about building capacity— among researchers and policy people—to produce and apply knowledge for better development results. This kind of influence can take years, or even decades, to take effect or become apparent. But it is no less important for that" (Ibid: 21). Finally, it is worth stressing that, while academic research can be important for policy development, especially when it focuses on policy problems, other types of knowledge also inform the PMP. In his recent literature review, Jones (2009) distinguishes, in addition to research-based knowledge: process-based knowledge, meaning the knowledge generated in the process of conducting research or development programmes and projects (implementing, monitoring and evaluating impacts);3 participatory knowledge, stemming from the views and voice of citizens involved in policy spaces; and multiple sources and interdisciplinarity knowledge. The latter refers to the need for a holistic understanding of the policy process, from a transdisciplinary perspective and including such factors as moral principles and judgments. More broadly, Davies (2004) identifies the following factors that, besides research-based evidence, influence policy and practice: experience, expertise and judgment; resources; values; habit and tradition; lobbyists, pressure groups and consultants; and pragmatics and contingencies. Clearly, research-based knowledge is but one of many influences upon policy and practice: recent studies in the US and the UK (Rigby, 2005; Rich, 2005; and GSRU, 2007) show "the relatively low status of academic research amongst sources of evidence used by policy-makers" (European Commission, 2007: 11).

3

Integrating this type of knowledge in the PMP faces many difficulties: "These include the capacity requirements and time constraints placed on project staff, who may often be more focused on action rather than analysis, and the wider difficulties of embedding a new working culture and providing the right incentives for learning and accountability. Knowledge of this nature is also often relatively context specific, and faces barriers to being immediately more widely usable. Use may also be hindered by power structures in development agencies, with pressures for staff to filter, regulate and fit information into prevailing management processes and frameworks (Mosse, 2006). Moreover, information about ‘failed’ projects is rarely published." (Jones, Datta and Jones, 2009).

7 1.1.2

Some recent typologies of research to policy frameworks4

As was said, many models –normative and interpretative– have been developed to try to capture and reflect how research and policy interrelate. Each model presents its own conceptualization of the nexus, its own assumptions and its own strategy for making research matter in policy. Thus, each model contributed something, and various literature reviews 5 have tried to give account of this by building typologies based on common ideas that frame different approaches. The point of departure is usually the rational-linear approach (Box 1.3). Though excessively stylised and very poor in describing the PMP, this knowledge-driven view somehow represents a common imagery associated with the uptake of research in the PMP. Grindle and Thomas (1990: 1164) even observed that "[t]he linear model has led donor agencies to support substantial efforts to strengthen policy analysis in developing countries in the expectation that good analyses will translate into good decision making and this into good policy. Operationalizing this expectation has generally taken the form of technical assistance contracts to build capacity in planning and policy analysis in many sectors".

Box 1.3 - Rational-linear models: research feeds directly into the policy-making process Knowledgedriven model

Driver: Assumptions: Limitations: Problemsolving model

Driver: Assumptions: Limitations: 4

Inspired by the linear innovation model (basic research → application), this model pictures a linear path from a new idea or innovation to its uptake at the policy level. In S&T areas, this approach has been used to explain policy changes that did not fit the distributive model (i.e. elitist control of policy issues), for example when these changes responded to diffuse interests instead of to powerful interest groups. To cite a case, the air bag regulation issue in the US originally seemed to fit the distributive model –the auto industry interests delayed several years the application of safety regulations– but, taking a longer term view, it best fitted the knowledge-driven model: the air bag system was supported by credible experts and then effectively pushed by policy entrepreneurs to bring about changes in the auto industry regulation, in spite of conflicting interests. Cases originating in the SSH research field are unusual. One could imagine, for example, a reputed research group effectively promoting, at the macro policy level, a methodological innovation that offers higher accuracy for the measurement of some relevant variable of the national economy. Supply: new knowledge presses toward its development and use. Production moves toward application in a linear and rational way. Simplistic view of the PMP, though it sometimes helps as a starting point to then turn to more realistic models, with additional variables and actors. Based on a stages' view of the PMP (problem identification; policy options examination; decision taking; policy implementation; monitoring and evaluation; and feedback into new problem or issue), this model considers that research can contribute to any of these stages through empirical evidence. The emphasis of the model is on expertise leading to diagnoses as inputs for decision-taking. Mode 1: SSH research enters the PMP through pre-existent information (searched for or acknowledged through any form of dissemination). Mode 2: SSH research is commissioned to fill knowledge gaps, shedding light on specific policy issues. Demand. Decision' needs lead to looking for research and applying it. Knowledge is a product. Centered on instrumental use of research. Knowledge is treated as 'data' and 'neutral' inputs. Optimistic

Many of the references cited in this section are from researchers who have worked in long term programmes on the knowledge-policy nexus, such as: Overseas Institute for Development (ODI): Research and Policy in Development (RAPID) programme, UK (www.odi.org.uk/rapid; www.ebpdn.org); Research Unit for Research Utilisation (RURU): a research partnership between members of the Universities of Edinburgh, Dundee, and St Andrews, UK (www.ruru.ac.uk); International Development Research Centre (IDRC): Evaluation Unit, www.idrc.ca/en/ev-26266-201-1-DO_TOPIC.html), Canada; UNESCO: Management of Social Transformationos (MOST) Program, Paris; Project “Knowledge and Policy” (www.knowandpol.eu); and Global Development Network (www.gdnet.org). 5 Among others: Jones (2009), Nutley et al (2007), UNESCO (2007), Lindquist (2001), Neilson (2001), Stone, Maxwell and Keating (2001), Keeley and Scoones (1999), Sutton (1999). In what follows, we do not pretend to give full account of the literature; rather, we provide general trends of the typologies developed by scholars who have been strongly involved in long term empirical work on the research-policy link.

8

Prescription to increase research use:

and simplistic view of research feeding the PMP, through discrete, predictable and manageable stages. Mode 1: improving the efficiency of communication links (effective packaging of research findings). Make good use of information technologies. Mode 2: improving terms of reference of the studies and the monitoring process (non-use is associated with poor quality of research).

Sources: Weiss (1979), Nutley (2007), Best et al (2010), 123HelpMe.com, 2010.

Based on slightly varying classification criteria, all recent typologies show how modelling the research-policy nexus has evolved from the rational-linear view to more comprehensive interpretations, derived mainly from advances in understanding the nature of policy processes, as well as from the application of social network theory, systems analysis, and even complexity theory. Box 1.4 offers a first grasp of this evolution based on Nutley et al (2007) typology, which differentiates between: traditional models, interactive/relationship models, and postmodern currents of thoughts. Next, we turn to two other recent typologies, and then complement the review with some insights and specificities related to the Southern context. Box 1.4 –Nutley et al (2007) typology of paradigms on the research-policy nexus TRADITIONAL MODELS In addition to the rational-linear models (already treated in table 1.1), this category includes mainly incremental models, the garbage can model, and policy streams models. Kingdom (1984) policy streams approach is well-known. It focuses on the agenda-setting process that shapes the PMP. Agenda and policy changes depend on the matching of different sets of factors and circumstances: "1) The problem stream denotes which issues are recognised as significant social problems. Citizens, groups and journalists work actively within this stream to trigger interest in problems. 2) The policy stream refers to which advice is regarded as ‘good advice’. This changes in tandem with the problem stream and external events. 3) The political stream: both the problem stream and the policy stream operate within a political environment characterised by elections, changes in government, changes in political champion causes, and changes in public opinion. 4) Policy windows occur when there is an opening for new views. This is usually triggered by a major event such as a crisis, a new international agreement, budget negotiations, a priority setting exercise, pressure group campaigns, election results, etc. Policy windows provide the opportunity to have alternative issues and solutions considered seriously" (de Vibe, 2002: 32). Nutley et al argue that "[p]olicy windows open relatively rarely, and do not stay open for long. At such times, policy entrepreneurs –individuals advocating certain policy ideas or proposals– play a key role in coupling solutions to problems and political opportunities. Policy entrepreneurs represent an important route for research to enter policy, for example by championing a set of findings that supports their position (Neilson, 2001)" (Nutley et al, 2007: 97). INTERACTIVE/ RELATIONSHIP MODELS This category comprises models that focus on communication and interaction as the key to the policy use of research. These models "emphasise the ways in which such knowledge will be negotiated, adapted and reconstructed within the contexts of its use, in a process that integrates research knowledge with local 'craft' knowledge. The emphasis is on joint processes of knowledge production" (Nutley et al, 2007: 123). Two communities models Caplan (1979) interpreted the non-use of research as a consequence of the cultural gap between the academic and the policymakers' communities. Researchers (scientists/experts) and policymakers (politicians/administrators/appointed officials) "live in separate worlds, with different and often conflicting values, different rewards systems, and different languages" (Ibid: 459). Thus, the limited use of research in the policy process is due to communication difficulties. More and better contact may improve understanding and research uptake, but effective interaction also involves value and ideological dimensions, in addition to the technical ones (Ibid: 461). Booth (1988) stressed on the differing incentives systems: "[the] structure of incentives within the academic community has also driven a wedge between social scientists and policymakers. These incentives attach greater weight to knowledgebuilding as against policy-forming research; to authoritativeness rather than usefulness; to the pursuit of rigor as against relevance; to the values of scientific independence as against the virtues of policy involvement; and to understanding rather than action (Ibid: 226). Wingens (1990) referred to a functional gap rather than a cultural one between researchers and policymakers. Because of their belonging to different social (sub)systems, in order to be adopted by the political system research needs to be adapted, recreated and transformed. "This model thereby integrates constructionist theories of learning, which propose that any new knowledge will be filtered through and shaped by pre-existing frameworks and experiences in the process of its use" (Nutley, 2007: 116).

9

Linkage and exchange model Lomas (2000) considered research and policy as processes. Its model (see figure below) divides the decision-making world into three interrelating domains: 1) Institutional structure (formal and informal) for decision making; 2) Values; and 3) Information. "Researchers who ignore the distinction between rational and sensible decisions, i.e. fail to acknowledge the influence of these political and institutional factors, are restricting themselves to a very limited niche in the decisionmaking world. A better understanding by researchers of the competing sources of information, the likely manner in which their findings will be purveyed into common knowledge, the nature of the decision-making structure/s, and the prevalent values will help them to know not only whether, but also how and when their findings might be useful" (Ibid: 144). The Canadian Health Services Research Foundation model (CHSRF, 2000) added two key groups in the PMP: research funders and knowledge purveyors. The model pictures the (ideal) relations between these two groups, researchers and decision-takers as follows: "Policy makers ask researchers about pressing problems, and researchers aim to supply policy makers with appropriate solutions. Research funders consult with policy makers around key problems, issues and priorities, and then translate these into funded programmes for research. Finally, through knowledge purveyors such as think tanks, conferences, journals and the media, findings from research (together with other forms of evidence) become ideas, best practices and interventions to be fed directly to policy makers. This model suggests that research use will happen when the links between all four groups are both mutual and strong (CHSRF, 2000; Lomas, 2000). It represents a 'virtuous cycle', in which any weak link in these relationships may inhibit the uptake of research within the policy community" (Nutley et al, 2007: 101).

Source: reproduced from Lomas, 2000: 143.

Context, evidence and link model The ODI-RAPID framework (Crewe and Young, 2002) is based on the study of 30 theoretical models (see figure below) that all contribute to answer the following question: Why are some of the ideas that circulate in the research/policy networks picked up and acted on, while others are ignored and disappear? The key variables that shape policy uses of research were identified and grouped into three overlapping categories (Jones, Datta and Jones, 2009): 1) The political context includes a range of factors, such as: the nature of the political system (e.g. authoritarian or democratic) and the level of democratic competition; the strength of government leadership; the relative strength of interest groups; incentive structures within policy-making organisations; capacities of both policymakers and institutions; and economic structures and processes. Jointly, these factors shape who is able to participate in the policy process, on what terms and how the process is structured. 2) The evidence: credibility, methods, relevance, use, how the message is packaged and communicated. 3) The links between PM and research communities (networks, relationships, power, competing discourses, trust, etc.). External factors, i.e., factors outside the specific context (socio-economic and cultural influences, donor policies, etc.) set up a fourth category of variables. For example, dependence on international donors and international financial institutions can enforce changes in policy content and processes, affecting the autonomy of national policymakers. The model represents the use of research as a dynamic, complex and mediated process, which is shaped by formal and

10 informal structures, by multiple actors and bodies of knowledge, and by the relationships and play of politics and power that run through the wider political context" (Nutley et al, 2007: 111).

POLICY NETWORK APPROACHES This view focuses on the role of networks in shaping research and policy connections. These types of networks include: 1) policy communities (specialists in a policy area, from within and outside government, who develop shared perspectives on a policy problem); 2) advocacy coalitions (groups of policy actors, who share policy beliefs within a particular policy sector, interact and form policy subsystems, and develop on the basis of shared normative and causal beliefs); 3) epistemic communities (networks of experts in a particular domain and an authorative claim to policy-relevant knowledge within that domain); and 4) issue networks (unstable membership networks, bounded around a given issue in a way that increases their interest and knowledge in that issue through policy debates and the like, rather than from their technical proficiency). POSTMODERN CURRENT OF THOUGHTS Postmodern approaches focus on the role of power: research is socially constructed through relations of power. In this case, the world of research and the world of policy-making are not separated; research does not get 'into' policy. "Postmodern accounts … pay attention to the processes through which different forms of knowledge are defined and accepted as legitimate in different contexts" (Nutley, 2007: 123). These approaches stress, among others, that knowledge only makes sense within the local context of its production. Thus, local research is important in the research use process, as well as experiential and tacit forms of knowledge held by policymakers (Ibid: 120-122). Source: Nutley et al (2007).

Jones’ typology (2009) contributes to a better understanding of power in the knowledge-policy nexus. The author differentiates three basic paradigms. The first one, rationalism, corresponds to the linear-rational view. The second, pluralism and opportunism, recognises that the policy process involves multiple actors and factors, uncertainty and pragmatic decisions: "The incorporation of knowledge involves often erratic and opportunistic processes, and explicit efforts of various actors" (Ibid: 5). The third paradigm, politics and legitimisation, is based on the view that "power is infused

11 throughout the knowledge process, from generation to uptake. Knowledge will often reflect and sustain existing power structures, and is used in the policy process in processes of contest, negotiation, legitimisation and marginalisation" (Ibid). The three approaches make different assumptions about the nature of knowledge and of policy: in rational models both knowledge and policy processes are assumed to be 'good'; the pluralist-opportunist models problematise the policy process but retain the assumption that knowledge is 'good' 6; and the politics and legitimisation view problematise both the nature of knowledge as well as the functioning of the process. According to Jones, post-2004 theoretical developments are generally based on the third paradigm (politics and legitimisation) and focus on power in three overlapping fields: actors and networks, institutions, and discourse. In the first field, knowledge is often considered in its tactical and strategic roles, subordinated to policy interests: the prevailing knowledge reflects the prevailing interests and the dominant networks (Ibid: 12)7. However, knowledge together with beliefs and values also play a role in motivating the development of new networks that, eventually, will carry information and ideas into the policy process. As was already presented in Box 1.4, networks models include issue networks, epistemic communities, policy communities and advocacy coalitions, and they illustrate different aspects of the flow of knowledge between actors in the policy process. The second field of analysis concerns institutions, which shape the formal and informal rules of the game, and affect which ideas and whose knowledge is used in the policy process: "… the political rules of conduct, … the political governance structures, … the governance processes, … the industrial relations regime, … set the parameters of what people talk about as well as who talks to whom in the process of policy making" (Schmidt and Radaelli 2004, in Jones, 2006). In this view knowledge is often altered and translated to fit institutions; it can become institutionalised or embedded in bureaucratic procedures, laws, programmes, departments, etc.; and it shapes actors' strategies and the political behaviour of bureaucrats. Institutions also influence the knowledge-policy link by orientating knowledge generation through providing funding for research in line with priorities or by commissioning research on given policy problems. Finally, the discourse analysis focuses on the frames and ideas PM use to select, organise and interpret information, as expressed in policy discourse.8 Language or discourse shapes the policy agenda, and causes problems and solutions to be understood or perceived in a certain way (Stone et al, 2001). Many scholars have studied how cognitive paradigms and policy narratives, among others types of ideas, influence the PMP, not necessarily for the best. It has been found, for example, that the use of cognitive paradigms can limit the range of options that policymakers perceive as useful9. Policy narratives might be very influential in informing policy-making and they are usually persistent, which means that new ideas often need to fit within existing narratives, or be very convincing, in order to feed the system. Jones notes that 'discourse' approaches on the knowledge-power link in 6

According to Jones (2009: 11), the concept of innovation systems falls in this category: "… work on innovation systems argues for the importance of both supply and demand of knowledge, the need for intermediaries and regulatory framework conditions (Rath and Barnett 2006), but retains an assumption that innovation and the uptake of knowledge will generally be ‘good’, that promoting such innovation will lead to social and primarily economic benefits". 7 Jones notes that "these dynamics are frequently very relevant for the link between knowledge and policy in the South. For example, recent work shows that the strength of economic interests plays a key role in shaping the policy process on a particular sector or issue, determining the extent and nature of evidence-informed policy dialogue (Pomares and Jones 2009)" (Jones: 12). 8 'Discourse' refers to the concepts and ideas that are relevant for policy, as well as the processes of communication and policy formulation that serve to generate and disseminate these ideas (Schmidt and Radaelli, 2004). 9 Rao and Woolcock (2007, in Jones, 2009) argue that a ‘disciplinary monopoly’ of economists at the World Bank has dominated much of the development work and has restricted what and how issues are studied, which in turn limits policy options and preferred strategies.

12 the PMP have been limited in their application, and that the question of how to redress power imbalances in the knowledge-policy link remains untackled. Among others, more in depth analysis are needed to comprehend why and how certain ideas are adopted as the dominant thinking in international development policy-making bodies (Perkin and Court, 2005; Jones, 2009). In the same way but at the national level, the question of how to substitute the use of narratives and models in a top-down approach in favour of interventions adapted to local contexts has received insufficient attention (Jones, 2009: 30). A very recent typology is that of Best and Holmes (2010), which focuses more broadly on knowledge than strictly on research. It considers three generations of thinking on the 'knowledge to action cycle' (KTA)(linear, relationships, and systems’ models) and, interestingly, the authors highlight the conditions in which each type might be an adequate framework of analysis (Box 1.5). In view of the complexity of the environment in which KTA takes place, they suggest to build analysis with a new systems’ lens: one that takes into account key dimensions of complexity theory thinking10 (Box 1.6).

Box 1.5 – Conditions for applying different knowledge-to-action (KTA) frameworks LINEAR MODELS Perspective: Knowledge is viewed as a product, generalisable across contexts and sectors. KTA is a pathway of discrete, predictable and manageable stages. One-way exchange process, from researcher producer to research user. Effective communication and 'packaging' of research findings is the key to research 'transfer'. Conditions for applicability: Classic dissemination and diffusion criteria are well-met (clear relative advantage of proposal, low complexity of issue, low risks and costs involved). Strong institutional structure and resources support the KTA process. Supportive culture and incentives for behaviour change at the practice/implementation level. RELATIONSHIP MODELS Perspective: Knowledge creation and knowledge use involve a clear commitment to close collaboration. Core processes for KTA are: linkages (partnerships, common interest's networks, etc.), collaboration and shared learning. Knowledge is 'exchanged'. Conditions for applicability: Consensual view among actors that local context and knowledge must be taken into account in adapting evidence-informed intervention strategies and instruments. Organisational culture favours evidence-informed planning, decision making and resource allocation. The complexity of the issue requires systems change to support change at the action level, and this is accepted by opinion leaders and decision makers. Stable research agenda and platform to support two-way communication and collaboration. SYSTEMS MODELS Perspective: KTA is a complex, dynamic, adaptive system, thus constantly changing. Culture, structures, priorities and capacities shape the system, which is made of various (sub)systems that exist within other interdependent systems. Key factors for understanding the KTA system are: the roles and actions of stakeholders, who are shaped by and in turn shape the system; and feedback loops during the process. The system requires activation to link its various parts. Conditions for applicability: All relevant stakeholders can be active collaborators in the modeling and solution seeking process; they are willing to invest time and resources to develop the model. Lead organisations view KTA as a key 'business strategy' that allows to integrate the model with the help of organisational change strategy. Source: Best and Holmes (2010).

10

Complexity theory is the study of interdependent, dynamic living systems and draws on work on ecosystem, maths, physics, and artificial intelligence, as well as a growing body of work on social, political and economic phenomena (Ramalingam and Jones, 2008).

13

Box 1.6 – On systems and complexity thinking The following is reproduced from Ramalingam and Jones (2008: 5, 8): "Systems thinking is particularly close in its origins and scope to complexity science. … complexity can only emerge in the context of a system, and certain aspects of complexity, such as feedback, find clear parallels in systems thinking (…).

Box 1: Similarities and differences between complexity and systems approaches It is worth noting that: • Systems thinking assumes that systems have dominant rules that can be used to calculate potential equilibrium, whereas complexity emphasises that systems tend to defy calculated equilibrium. • Systems thinking sees that systems have some kind of ‘control system’ that provides guidance and shapes the system, whereas complexity recognises the possibility of self-organisation. • Systems thinking suggests that elements in a system can be understood as isolated elements and symbols, whereas complexity forces us to see the interdependence of the nature/meaning of individual elements and the context in which they are embedded. • Systems thinking assumes that systems propose rational processes and predictable results, albeit through complicated means, whereas complexity recognises that solutions are arrived at via dynamic processes that are not likely to result in a final conclusion. • Systems thinking assumes that systems change their structures in accordance with rule-based learning, whereas complexity recognises that change is perpetual, so learning is a constant factor. [However s]ome thinkers have gone so far as to outline the key differences – see Box 1 below. (…) The three sets of complexity science concepts are as follows: Complexity and systems: These first three concepts relate to the features of systems which can be described as complex: 1. Systems characterised by interconnected and interdependent elements and dimensions are a key starting point for understanding complexity science. 2. Feedback processes crucially shape how change happens within a complex system. 3. Emergence describes how the behaviour of systems emerges –often unpredictably– from the interaction of the parts, such that the whole is different to the sum of the parts. Complexity and change: The next four concepts relate to phenomena through which complexity manifests itself: 4. Within complex systems, relationships between dimensions are frequently nonlinear, i.e., when change happens, it is frequently disproportionate and unpredictable. 5. Sensitivity to initial conditions highlights how small differences in the initial state of a system can lead to massive differences later; butterfly effects and bifurcations are two ways in which complex systems can change drastically over time. 6. Phase space helps to build a picture of the dimensions of a system, and how they change over time. This enables understanding of how systems move and evolve over time. 7. Chaos and edge of chaos describe the order underlying the seemingly random behaviours exhibited by certain complex systems. Complexity and agency: The final three concepts relate to the notion of adaptive agents, and how their behaviours are manifested in complex systems: 8. Adaptive agents react to the system and to each other, leading to a number of phenomena. 9. Self-organisation characterises a particular form of emergent property that can occur in systems of adaptive agents. 10. Co-evolution describes how, within a system of adaptive agents, co-evolution occurs, such that the overall system and the agents within it evolve together, or co-evolve, over time."

In a recent sociology-based approach to the knowledge-policy relation, Delvaux and Miguez (2008)11 construct a tentative theoretical frame, within the emergence of the knowledge economy (Box 1.7). They distinguish two dimensions: that of the policies informed by knowledge; and that of the policies governing knowledge. The model is illustrative of the sophistication of theoretical developments in 11

This work was developed in the frame of the 2006-2011 KNOW&POL (Knowledge and policy) project (www.knowandpol.eu/), of the Sixth RTD Framework Programme, funded by the European Commission. It involves 12 research teams analysing how different sources of knowledge are mobilised in decision-making in education and health in six European countries.

14 the North in this regard, and is also revealing of the limitations derived from looking strictly at the 'research'-policy link.

Box 1.7 - Towards a sociology of the knowledge-policy relation (Delvaux and Mangez, 2008)  Knowledge is defined in a broad sense, including both scholarly and lay knowledge, relatively stable knowledge rooted in belief, and knowledge 'put into language'. Knowledge not only circulates but also never ceases to reconstruct itself as it circulates: "Everywhere, actors are interpreting the world according to their beliefs; everywhere, they are mobilising, combining and circulating knowledge put into language, especially when exchanging multiple arguments in the course of controversies, or in developing proposals. In so doing, they produce new knowledge, evaluated with public action in mind, and whose novelty resides less in the nature of its various constituent elements than in the way it is combined and adapted" (Ibid: 61).  Public policy is considered as public action, i.e., the outcome of actions occurring in multiple 'scenes' (settings) that interact in horizontal rather than vertical fashion, and in a circular rather than linear fashion. Public action develops within a structural framework, but is able to modify the factors that define its parameters. It is thus fragmentary and flexible, an d its course unpredictable. The framework is composed of four elements that are closely linked. On the one hand, beliefs (i.e., normative and cognitive frameworks to which the actors refer to interpret the world, justify their actions or guide their behaviour and practices), and knowledge circuits (knowledge flows circulating among actors, scènes, sectors or countries), which both structure knowledge. On the other, forms of coordination (formal and informal mechanisms and rules defining the parameters of social interaction) that regulate the scene and its environment, and configurations of interdependence (the structure of the positions occupied by different actors, scènes or sectors relative to one another), which both structure social interaction. These four elements structure all scènes. None of them is pre-eminent, and each tends to adapt to the other factors when public actions address one of the factors.  "Control over what will develop in the 'target scène' involves knowing how these four factors shape it. It also assumes having the resources to affect at least one of these factors. Influencing beliefs, the first factor cited, is not conceivable in the short term, regardless of which actor takes the initiative. Modifying formal institutions is easier for those who have the power to legislate or regulate; however, it is not out of reach to other actors, even though it might be indirect -through pressure on actors who have this legislative or regulatory power. Circulating knowledge is more within the reach of the various scènes, and is increasingly playing a role as a form of action. Lastly, influencing the position that the scène and its actors occupy in the interdependence structure involves either (a) being able to give to or withdraw from this scène, or from certain of its actors, the resources they desire, or (b) influencing the scènes and actors that have these sought-after resources in order to release them or hold on to them. As research on the implementation of public policy amply demonstrates, the intervention of a scène in one or another of the factors structuring a ‘target’ scène and its actors has uncertain effects" (Ibid: 23).  In each scène every action or decision involves two prior processes that are particularly knowledge-intensive. The first –problematisation (proposing problems)– transforms situations into problems previously considered as non-problematical or inevitable. The second –idea creation– is the process in which proposals and alternatives emerge and are selected. (In conventional phrasing, this would be equivalent to the processes of agenda setting and policy design.) The authors draw on Kingdon streams model (see Box 1.x) to explain the processes of defining problems (problem stream) and fabricating ideas (policy or ideas stream), and agree with Kingdon that these processes are not sequential (the former does not necessarily precede the latter) neither rational. However, Delvaux and Mangez introduce the assumption that problems and ideas do not emerge or survive unless they have passed certain tests (naming, importance, accessibility, prioritisation, compatibility, feasibility, etc.), which vary for each of both processes and call for the mobilisation of different types of knowledge. Much of their study is dedicated to expose this assumption or how these tests (filters) operate.  The 'mobilisation' of knowledge (embedded in beliefs, scholar or lay, narratives, etc.) plays a particularly important role in these two essential public action processes: the definition of problems and the fabrication of ideas. The knowledge mobilised in each scene depends particularly on the forms of coordination prevailing in the scène (hierarchical, formal, democratic, deliberative) and on the position of the actors and the scène in the knowledge circuits. External actors (to the scenes) also matter: some are able to act specifically on the knowledge mobilised by actors directly involved in the scène; and their position in the knowledge circuits allows them to have control over the knowledge circulating in or regarding the scène involved. Controlling the circulation of knowledge then becomes an issue of the interactions between actors and scenes. However, "[t]he interaction among scènes unfolds against a backdrop of interdependencies that are, to be sure, often asymmetric yet make each scène dependent on other scènes" (Ibid: 24). Naturally, the richness of this framework derives less from the model itself (oversimplified here above) than from giving grounds for an extensive analysis on: the links between the numerous ‘scènes’ that formulate public action; the factors structuring the latter; and the role of knowledge in defining problems, fabricating ideas and providing an instrument of power. Source: Delvaux and Mangez, 2008.

15 Finally, we close this section with some very recent reflections on the bottlenecks that threaten optimal trajectories between the realm of politics, policy-making and useful research (Box 1.8).

Box 1.8 - The bottlenecks that threaten optimal trajectories between the realm of politics, policy-making and useful research The issue of these bottlenecks are the backbone of several policy reflections. In the realm of “knowledge democracy”, a term under which an international conference was held in The Netherlands in 2009, different questions in this regard were posed within the context of scientific research and innovation (‘t Velt, 2010: 1-10), being thus particularly useful for our purposes. Some of the answers given are:  “The actual political agenda may not correspond with the existing policy theories that are either laid down in existing policies, legal systems budgeting rules, etc., or/and are embraced by the top civil servants” (Velt, 2010: 9). Comment: the policy theories embraced by the top civil servants are issues that deserve close attention. Nelson and Winter reflect on this in the following way: “Just as voters sentiment generally provide only loose constraints on the actions of elected officials, so the decisions of elected officials generally leave considerable amount of discretion to the civil servants and others who carry out a program or policy. Prior to the 1960s the role of ‘administration’ was seen in the political science literature as simply technical, consisting of working out the best way to achieve an objective or carry out a policy defined by elected officials and mandated by the electorate. Since that time, it has become better recognized that the shape of a policy is to a considerable extent determined by how it is implemented” (Nelson and Winter, 1982: 377).  “The translation of policy questions in knowledge demand may prove to be extremely difficult, for instance because the policy objectives bear a symbolic character, or because the policy questions are wicked in nature, lacking underlying consensus on values” (Velt, 2010: 9). Comment: the fuzziness that may affect knowledge demand on the part of policy makers can be related to the fact that what policy makers want to accomplish, associated to the decisions they have to make, can be unclear even to themselves. Sfez, making the critique of decision-making, points to the following: “Decisions appear as a phenomenon that is produced in an open system, under certain conditions: contrary as what is usually thought, such decision phenomenon is not produced in places of nuclear, instituted power, nor outside the systemic circles, in ‘outcast’ places. Decisions are neither taken in those places were perfect and univocal information is exchanged, but there were a truncated, translated and deformed information pass from one subsystem to another and then can be transmitted and put into practice.” (Sfez, 1986: 249)  “Inconvenient truth, newly produced knowledge that attacks the existing policy theories, will probably not be applied in policy-making” (Velt, 2010: 9). Comment: one of the interviewees in our study expressed exactly this, even with similar wording, referring to the policy reception that research on some biotechnological issues had from policy makers. It can be added that when research analyses on hot policy issues, for instance the TRIPS-plus agreement and its possible effects, are presented to policy makers who have already decided to sign such agreements, its fate will probably be to be disregarded, even with contempt, by policy makers.  “Research will produce knowledge in the future but the need is urgent, and the political agenda is slightly volatile so there is a general problem of timeliness. In order to recognize the time lags just described on the one hand and the legitimate demand for useful new knowledge on the other we should attempt to design the policy agenda in the near future instead of the present one, but that is a dangerous activity”. (Velt, 2010: 9).

1.1.2

Studies focusing on developing countries

Looking at studies carried out to get a better understanding of the research-policy nexus in developing countries, it seems that empirical research has tended to be highly influenced by the rational-linear models. The relevance of different paradigms, the role of different types and sources of knowledge, and the role of different actors in the PMP are questions that should be analysed from an 'open-ended stance' in the South (Jones, 2009). Some recently published work present the outcomes of projects addressing these issues, though. Based on an in-depth analysis of IDRC founded research projects, Carden (2009) found that the following distinctive features affect research uptake in the South with respect to the North: Democratic institutions and customs are often more precarious. International financial/aid institutions are often considerably influential in policy-making.

16 Policy design and implementation challenges are (even) greater (weak policy design capabilities; inadequate administrative, legal, or management capacity; implementation incompetence unchecked by sufficient monitoring and accountability). Demand for research can be missing: "much of the Northern-based knowledge-to-policy literature assumes an active demand among policymakers for the knowledge that research can supply. Hard-pressed decision makers in governments of developing countries often know little about the help that research can offer them, and are therefore indifferent to the value of building local research capacity in the long run". Policy networks and decision regimes that facilitate the demand and supply relations between decision-makers and researchers do not necessarily exist in the South. Developing countries often lack the intermediary institutions (knowledge brokers) that carry research to policy. Staff turnovers in research organizations and in government tend to be higher. Researchers often lack hard data to drawing reliable conclusions Policymakers lack confidence in their own researchers and often turn to foreign experts. Research findings and methods usually travel North to South. Research experiences in the South are often neglected, entailing a loss of opportunities for information exchange among Southern researchers. ‘Personal’ relationships can lead to misgovernment. Another large scale study on developing countries has been carried out by the research team of the ODI-RAPID programme and a number of partners (Jones, Jones and Walsh, 2008). The analysis combines theoretical and empirical work, and draws on the view of 600 developing and developed country stakeholders, key informant interviews with 30 global experts, and six country case studies. It focuses on three questions: “What is the patterning of relationships between scientific researchers, policy decision-makers and intermediaries in developing country contexts? What are the challenges and opportunities for strengthening these linkages? What types of strategies exist or could potentially be adopted to improve evidence-informed policy processes?" (Ibid: vii). The authors hold that knowledge from the natural sciences faces similar barriers to knowledge produced by other disciplines in terms of its uptake in policy processes: a lack of political will; poor communication between researchers and policy-makers; limited responsiveness of research to current policy concerns; ineffective lobbying and inappropriate targeting; and limited researcher credibility in the eyes of policy-makers. They then highlight a number of key tensions and challenges in incorporating STI-related knowledge effectively into policy dialogue in developing countries (Box 1.9).

Box 1.9 – Improving S&T research uptake in the policy-making process in developing countries The following is reproduced from Jones, Jones and Walsh (2008: viii-ix):  Poorly institutionalised evidence-based policy-making needs to be recognised and tackled. Evidence-based policymaking is poorly institutionalised as a process in developing country contexts. Owing to a lack of accountability and/or formal mechanisms for the integration of scientific knowledge into policy, scientific research is often used selectively at the discretion of policy-makers. Therefore, policy priorities often drive the usage of research, rather than research stimulating policy recommendations. Knowledge translators and knowledge brokers need to be mindful of this when developing strategies to communicate scientific, technological and innovation (ST&I) research findings to policy audiences.  Audience-appropriate information targeting is imperative. ST&I information must be targeted according to the needs of specific actors in the policy process and the stage in the policy process at which different actors use ST&I information. Science-oriented ministries were primarily interested in science for agenda setting and policy formulation, whereas nonscience ministries relied more on scientific research at the policy implementation and evaluation stages.  Intermediary organisations are needed to act as knowledge brokers and capacity-builders for researcher and policymaking communities. Scientific researchers often conduct research in line with long-term goals, whereas policy-makers require information that responds to short-term goals. Researchers tend to use technical jargon and embrace uncertainty

17 and risk, whereas politicians desire language that is policy-relevant. Accountability lines differ as well, with researchers answering to funders, and policy-makers to their constituencies, stakeholders and political parties. There is therefore a strong need for intermediary organisations to act both as knowledge brokers at the science–development policy interface and as capacity-builders for both researchers and policy-makers. However, the degree of unmet need is such that few successful pathways forward have been identified. This suggests that various options could usefully be piloted and evaluated.  Interaction and deliberation rather than uni-directional research dissemination is needed to bridge the ST&I researcher policy-maker gap. In order to bridge the gap between ST&I research and policy-making, there is a strong need to go beyond the dissemination of research findings. Instead, greater interaction, discussion and deliberation between researchers and policy-makers are called for. While online formats were considered useful, face-to-face interactions were preferred by most policy-makers.  Policy-engaged scientists are critically important. The complexities of the policy environment often intimidate researchers, owing to the risks of the politicisation of science as well as a limited understanding of the policy-making process. However, there is a strong desire on the part of Southern policy-makers in particular for scientific findings to be complemented by policy-relevant recommendations. Policy-makers and development practitioners would be able to make greater use of scientific research findings if scientists would engage more openly with the resulting policy implications and present a range of possible policy options. Particularly when government priorities lie in social science areas, such as poverty reduction, the relevance of scientific information for development policy must be communicated.

 Improving public understanding of ST&I will facilitate better policy dialogues.There is a strong interest by policymakers and researchers in greater public participation in ST&Irelated policy debates, facilitated by initiatives to improve public understanding of ST&I so as to promote the emergence of a more informed and engaged public. However, although fears about the risks of the democratisation of scientific knowledge identified in the theoretical literature were largely not borne out by developing policy actors active at the science–policy interface, the challenges involved in reconciling Western scientific knowledge and indigenous conceptualisations of knowledge were recognised as considerable.

A last issue of interest we put forward because it has been little studied, is whether the use of research depends on the nature of the sector in which influencing is intended. Jones, Datta and Jones (2009: 1) argue that "[k]nowledge–policy dynamics differ across policy sectors due to divergent actors, demands for new knowledge, and capacities to use such knowledge. Some sectors, like trade, require highly technical expertise, while others, like education and natural resource management, involve increasingly extensive consultation processes. Vested economic interests might play more of a behind-the-scenes role in certain sectors, as might international debates. More contested sectors might also find less room for evidence …". Although, again, this does not concern exclusively SSH research, it is noteworthy to consider the framework that was developed to examine possible differences according to two sets of variables related to the policy issue in question and to the broader policy process (Box 1.10). Box 1.10 – Varying factors across different sectors/issues in the research-policy dynamic

VARIABLES OF POLICY ISSUES/SECTORS  Type and level of technical expertise required in a particular sector: a higher level entails increased demand for knowledge and thus more room for research uptake. However high expertise requirements can result in a single discipline monopoly and restricted dialogue.  Level of contestation: a high level might hinder research uptake in favor of other factors (emotional, religious, etc.)

VARIABLES OF POLICY PROCESSES 

Number, density and strength of issue champions: the number of (collective) actors and their effectiveness in advocating or defending an issue are relevant variables in policy changes.



Stage of the policy process at which a policy issue is raised: knowledge needs and use vary at different stages in the policy process.



Policy change objective sought: incremental changes can be reached through an instrumental use of research but paradigmatic shifts require new and challenging policy discourses.



Institutional capacities: higher institutional capacities both in the policy sphere and in knowledge institutions tend to facilitate research uptake.

 Strength of economic interests: when they are strong they may dominate policy debates, over other knowledge and actors.  Level of internationalisation: when epistemic communities and advocacy networks develop across countries, knowledge actors play a stronger role in policy-making if they are able to coordinate actions transnationally.

Source: Jones, Pomares and Pellini (2009).

18

1.2

The role of SSH research in innovation policy development

The previous section treated the research-policy nexus from a general SSH perspective. Before turning to our empirical work, we make a brief incursion into this nexus from a specific STI perspective. After all, what is sometimes termed the ‘innovation paradigm’,12 more often referred to with the concept of ‘National System of Innovation’ (NSI), has become a framework of thought commonly used in policy making in Latin America (LA). And most international organisations involved in STI development now rely on the NSI conceptual framework to justify the promotion of their policy instruments in the developing world. Thus, a relevant question to include as background of the present study concerns the role social sciences played in the transition from an approach to growth and competitiveness centrally based on R&D stimulation, to a focus on innovation. In this regard, Mytelka (2001) provides illustrative insights on the intertwined role played by, on the one hand, ‘dissident’ social scientists and, on the other, some international organisations in bridging the gap between innovation theory and innovation policy in the 1980s and the 1990s.13 His concluding remarks (reproduced in box 1. 11) also show some of the unsolved policy problems at the beginning of the 2000s, which in fact are still debated nowadays. Box 1.11 – Innovation theory and innovation policy: bridging the gap (Mytelka, 2001: 134-135) “Innovation theories emerged in a period of dramatic change. Expectations of growth were diminishing after several positive post-war decades. Technological ruptures were underway but their impact on productivity was not yet felt. Imports from low-wage countries were increasing and, coupled with new patterns of investment and organisational change, created further economic dislocation as regions declined and unemployment rose. Existing theory could not deal with these changes and the paradoxes to which they gave rise. While national governments in the developed world initially fell back upon neo-protectionist solutions and then embraced liberalisation, a small number of international organisations such as the OECD and the European Commission, became the locus for exploratory thinking around the issue of technological change. Dissenting theorists slowly reformulated the problem as one of learning and innovation and contextualised it in terms of innovations systems and institutions. Passage through international organisations then served to legitimise these concepts and to promote them as focusing devices in national policy making. (… ) In this process, and despite their “outsider” status, social scientists working within the new innovation paradigm have been extraordinarily successful in building a constituency for innovation systems approaches and in the design and redesign of innovation policies. By emphasising the contextually specific nature of innovation processes, they brought theory closer to policy, but have not entirely bridged the gap. Nor has the emphasis on a holistic and differentiated approach implicit in the innovation system literature made the task of its use in the development of policy instruments any easier. Evolutionary theory, for example, “would predict that different actors would do different things. They would see opportunities differently. They would rank differently those that all saw.” (Nelson, 1996, p. 125). We would thus expect national governments to tailor new policy instruments to the particular habits and practices of actors whose behaviour policy is designed to influence. Only where stakeholders at the regional level have been able to shape policies directly through participatory processes are there small signs of movement in this direction. For the most part, policy makers have been hard pressed to deal with the complex reality that innovation system approaches represent. The absence of a unified theory that relates innovation to growth and distribution and links macro-approaches to the micro level has slowed the application of innovation theory to policy areas beyond the narrow confines of education or research and technology development policy. Similarly, the lack of new measurement tools has limited the translation of innovation theory into effective policy instruments. (…) Concurrent developments to measure innovation have been undertaken in the 1990s. Paul David, Richard Nelson, Keith Smith and Luc Soete were among those who played a role in efforts at the OECD and in the EU to build an empirical base for the analysis of innovation… But these efforts have yet to provide the tools, for example, to test the OECD’s conceptually interesting hypothesis that a system’s innovative capacity is related to the extensiveness and efficiency with which it distributes and absorbs knowledge (David and Foray, 1995). As this chapter has shown, although innovation theory has made considerable conceptual inroads, there is still a way to go before the links between innovation and other policies are well established and the ability to measure the results becomes a reality.

12

Wether this is a paradigm in the Kuhnian sense is not a clear-cut issue (see Martin, 2009), but there certainly exists a shared body of work that originated in the pioneer work of scholars such as Freeman, Nelson, Pavitt and Rosenberg, whose contribution has been called the “Stanford-Yale-Sussex synthesis”. 13 Of course, tracing back the origin of innovation theories, Schumpeter would indeniably appear as ‘the’ pioneer in this field, highly dissident from the mainstream economics at the time of his writings (1910s-1950s).

19

As the cases of Keynes and Prebish already illustrated in earlier decades, the uptake of theory in policy-making depends not only on the quality of research but also on a particular opportunity that emerges from a combination of factors. In the case of innovation, a first factor was that economic crises showed the limits of the mainstream framework of thought and opened the way for dissenting views. Second, at that time, organisations such as OECD, the European Commission and UNCTAD were playing an important role in consensus-building on the way to tackle the issues of rapid technological change and globalisation; they were therefore more flexible and open to new views than stronger but more hierarchical organisations such as the IMF and the World Bank, more prone to stick to the neoclassical-based, macroeconomic perspective. Third, this search for new views was successful because there actually was a group of ‘outsiders’14 investigating on different aspects of innovation and willing to work closely with the former international organisations. This collaboration brought about changes in the research programs of these organisations and “to a learning process that ultimately led to a reformulation of the problem and to a reconceptualisation of the search for solutions” (Ibid: 126). The bridging between theory and policy thus implied crossing over the boundaries between academia and organisations. This illustration of research-policy interfaces in STI leads us to highlight the branch of social sciences that focuses on what matters for building effective policies in that area: STI policy research. It draws on a wide range of social science disciplines –mainly economics, sociology, political science, organisational science, business and management science, and psychology− and “rather than being theory-driven or paradigm-driven, it is primarily a problem-oriented field that focuses on practical issues to do with specific policies for science, technology and innovation, taking account of the central role of firms in the evolution of technology and innovation” (Morlacchi and Martin, 2009: 57215). It is a distinct research field of social sciences than science and technology studies (STS) and technology and innovation management (TIM):

Disciplinary composition of research community

STI POLICY Economists (of innovation) + sociologists, political scientists, historians of technology, and management and organisation scientists.

Units and levels of analysis

Firm, industry, and national levels

Main focus

Role of PM in regulating or facilitating market interactions and collective processes among firms and other organisations.

STS Sociologists (of S&T) + philosophers and historians of science

Processes of scientific research and technological innovation. How society shapes S&T, and how in turn S&T shape society and the environment.

TIM Business and management scientists + economists, administration and organisation scientists Firm and to a lesser extent industries/sectors How to best manage R&D and innovation.

Source: Based on Morlacchi and Martin, 2009

Morlacchi and Martin argue that there are four interacting components in STI policy research, which all should be addressed adequately to enable this research field to collectively and effectively fulfil the responsibility of “serv*ing+ the ends of society, helping to construct more effective policies for science, technology and innovation, which in turn will yield greater benefits for society” (Ibid: 572573): 14

These included industrial economists, economic historians, economic geographters, political economists and others on the margin of mainstream economics, such as Lundvall, Nelson, Freeman, Soete and others. 15 This section draws on Morlacchi and Martin’s editorial of a special issue of the Research Policy journal on: Emerging Challenges for STI Policy Research: A Reflexive Overview.

20 STI policy science: focuses on seeking the most technically correct answer to political problems in terms of available social scientific knowledge STI policy engineering: refers to the use of a set of procedures to determine the technically best course of action to implement a decision or achieve a goal STI policy entrepreneurship: is committed to searching opportunities for applying determined solutions. STI policy scholarship: is concerned with shaping ways of thinking and learning about society’s problems, and “to understand how key actors in the policy process come to understand those problems”. Viewed this way, and with regard to our previous discussion on the research uses, STI policy research thus assigns itself both instrumental and conceptual functions. Looking at the overall evolution of STI studies in the last 50 years, it is interesting to note a certain parallelism between the frameworks that developed to gain a better understanding of the innovation process and those that intended to model the SSH research-policy nexus. In both cases it started with linear, rather simple models (science-pushed and demand-pulled in the case of innovation) that were gradually refined and made more complex. Evolutionary economics provided a framework for the emergence of a systemic view of the innovation process (NSI and local systems of innovation) and, presently, ‘systems theory’ and ‘complex systems’ ideas start to be integrated in the STI policy discussion, in a similar way as we commented on the knowledge-policy nexus in the previous section. A recent paper of Aghion et al (2009) analyses the question of whether workable STI policies can be designed and evaluated in a “systems-theoric” framework. They state: “… we recognize the virtues of a systems approach to technical change and innovation. Such an approach helps to highlight and capture several characteristics of the process of innovation and technological change that are of direct relevance to technology policy. These characteristics involve: (i) the multi-directional links at the same point in time between the stages of technological change; (ii) the cumulative processes over time leading to feedbacks and lock-in effects; (iii) the dependence of technological change upon knowledge and the assimilation of information through learning; (iv) the unique character of the details of the development path and diffusion process for each innovation; and (v) the systemic and interdependent nature of the process of technological change.” The authors then develop some ideas on how a complex systems’ view could be used to deal with innovation, economy and growth. However, they conclude that “*t+he practical difficulties of designing ‘interventions’ for a system of such complexity pose formidable challenges because at least some among the conditions that call for government policy interventions also imply that important aspects of the system’s behaviour may be “emergent properties” that cannot be reliably deduced from a knowledge of the properties of its constituent parts. [We conclude] with a few cautionary reminders of the political hazards that await policy researchers and practitioners who suggest that their work on large and complex systems should be evaluated on the basis of observed policy “outcomes”.

21

2.

Empirical evidence on the research-policy nexus in STI

This chapter is based on information obtained through 50 interviews, conducted during 2009 and 2010, to researchers in SSH working on innovation and to PM involved in the design of STI policies. The first group includes 35 interviewees and the second, 15. The list of interviewees with their institutional adscription can be found in Annex 2. The country coverage in LA includes: Argentina (4 interviews), Brazil (6), Chile (4), Colombia (3), Costa Rica (2), Cuba (1), Mexico (8), Uruguay (7), and Venezuela (8). This was complemented with interviews to some researchers from developed countries: Denmark (1) , Spain (2) , Sweden (1), and United States (3). Interviewees were selected after we identified the main SSH research groups on innovation in selected LA countries. With the help of renowned researchers in different countries, we initially identified at least three researchers and three PM in each case, though the number of interviewed persons per country was finally uneven. Researchers were selected for their longstanding trajectory on innovation policy issues and, in general, for their leading function in an innovation research group. Interviews were carried out either face-to-face with the person, or through the Internet (SkypeTM). Some persons answered the questionnaire by e-mail. All interviews were transcripted and then transferred to a qualitative data analysis program (Atlas.ti). As the title of this document indicates, this is a perception study. This means that the main source of information for our analysis is thought and experience of a group of selected people, collected through a series of questions. Interviews, which usually lasted for more than an hour, were conducted with a semi-structured questionnaire that was applied flexibly enough to get an in-depth opinion of participants on the topics at stake. This way of proceeding also meant that sometimes some questions (or sub questions) had to be skipped due to time constraint. Also, in cases where the questionnaire was e-mailed, an ‘executive’ version of it was used. All this restricted the possibility of a quantitative processing of the answers but we clearly aimed at a qualitative analysis that would reflect a diversity of opinions rather than yes/no or multiple choice answers. The questionnaire addressed to researchers (available in Annex 3) focused mainly on the following topics: Origin, research interests and agenda fixing of the research group Perception of the influence and use of the group’s research Relations with other innovation research groups at the national and international levels Opinion on the innovation PMP in the country, its actors, inputs, etc. Bridging research and policy: obstacles and suggestion actions. The questionnaire to PM (available in Annex 3) addressed the following issues: Main features of the innovation PMP in the PM country (inputs taken into account, identification of innovation demands, articulation with other policies, etc.) Perception of the relevance of SSH research for innovation policy development Bridging research and policy: obstacles and suggestion actions. The next sections systematize the main results of the survey.

22

2.1

Brief characterization of SSH innovation research in LA countries

In this section, we present some features of the innovation research carried out in SSH in LA, based on the information provided in interviews: its origin, the groups’ research interests and their modes of selecting research projects. Origin Researchers pointed out that innovation studies are relatively recent in LA, in spite of the fact that Sabato and Botana developed as early as in 1968 a quite influential concept: the Sabato Triangle of relations between Government, Academy and Industry. In the early 1990s innovation became a main issue of inquiry for Latin American researchers, though STI studies carried out from a SSH perspective continue to be scarce today. Moreover, these studies lag in relation to other fields of work of SSH and have not given rise to a critical mass of scholars. - “In the 1980s practically nobody was dedicated to STI; very few groups were studying these issues systematically. I think that interests on the subject exploded somewhere between the late 1980s and the early 1990s (...) I think that the oldest post-graduate program is that of Campinas, organized by Amilcar Herrera around 1985. In 1988, we developed a master program at the University of Buenos Aires on Policy and Management of Science and Technology. (...) Today interest has risen, even if it is not particularly strong. I think that in relation to other fields of SSH, the social studies of S&T are relatively at the margin.”

Innovation research groups coincided in considering that a good deal of the blossoming of this research is the result of the growing conviction that innovation processes are a key to development. In this sense, studying innovation processes in Latin American countries has become relevant, as illustrated by the following comment: - “Having few innovation activities is a very important source of underdevelopment. I was always interested in inquiring around specialization profiles, issues of structural change, and this is why I entered the field of innovation.”

Simultaneously, the need for critically revising the theories that were developed in highly industrialized countries was acknowledged: - “There is a complex story that can be briefly told. Almost every theory passed from a very international and exogenous vision to a more endogenous vision. That is, we all come from worldeconomy ideas either of dependency or of neoclassical thinking, the later saying that if we let all elements of the international dynamics move freely (production factors, goods) this will diffuse strong development dynamics. But all theories have moved, even at ECLAC, looking for more endogenous elements, searching for explanations based on local institutions and actors.”

Interests guiding the innovation groups' research Researchers expressed a variety of interests that guide the research performed by their group or subgroups, namely: i. to generate STI information and build indicators ii. to study different modes of knowledge generation and its creative application, in the realm of production, health, or social inclusion iii. to understand development processes iv. to provide knowledge to the process of policy making v. to analyse the innovative dynamics at the microeconomic and sector levels, and modelling it vi. to analyse the public perception of S&T vii. viii.

to study impacts derived from innovation processes in specific settings, e.g., in the environment and the organization of work to analyse how contexts characterized by scarcity of resources influence research agendas.

23 Interests grouped in the first category refer mainly to statistical information and indicators. Thus, it includes improving ways of measuring innovation, systematizing innovation indicators, developing new tools to detect social impacts of innovation, and generating new indicators reflecting cooperation and networking intensities. Topics grouped in the second category cover: social technologies, innovation and social inclusion; history of the relations between scientific knowledge production and use; modes of knowledge production; and origin, organisation and evolution of scientific research groups. Interest in understanding different modes of knowledge production and the creative application of this knowledge is exemplified in the following: - “Innovation and social inclusion are issues with a normative turn. When one studies development issues, one goes into issues of equality because inequality is one of the marks of underdevelopment. When one studies equality, one finds that in fact there are two types of equality. One that can be termed distributive or reactive equality, and another one that can be termed pro-active equality, more related to fostering innovation while improving equality. The idea relates to different modes of modernization: some modes are 'exclusive', and others 'inclusive' or solidary. This is associated to an idea that links innovation with guiding, normative factors.”

Regarding the third category, Uruguayan researchers in particular expressed their concern for understanding development processes. Their starting point is that a developed society is defined by its capacities to make development happen. Understanding what these capacities are and where they lie is thus essential. Besides economic capacities, there are institutional, social, cultural and political capacities that are worth understanding in their interplay: - “The issue is to capture which are the determinants of these capacities for development, and then to understand how they develop. That is, which is the historical process that fosters the genesis of these capacities? To what extent are they universal or specific? If they are specific, how much does this specificity owe to historical and cultural reasons? A key issue concerns the relationships between knowledge and development in the South, and the characterization of these relationships. Another important issue is to understand how different underdevelopment is today given the new role of knowledge.”

Regarding the fourth category, practically all groups mentioned their interest in generating knowledge to contribute to the policy-making processes. This interest translates, e.g., into studying different models of science and technology policies, their historical evolution, and the stages of their transformation viewed from a social perspective that links these stages to development processes. In general terms, the research approach is that innovation is a social process and that the articulation between actors can foster or hamper innovation. Thus, institutional arrangements and policy formulation have been central in several research programs: “Innovation policies are confronted with scarce resources, but the problems such policies face do not only stem from a lack of finance. More and more problems are related to institutions, their channels of communications, their rules... Such rules are taken by actors involved in diverse processes associated to knowledge and innovation as norms of behaviour, and this can be a serious problem because sometimes these rules are associated to old schemes so that a new design of incentives, for instance, fails because people are too strictly attached to the old rules.”

Some groups have centred their attention on the entrepreneurial behaviour, particularly on the propensity to innovate exhibited by entrepreneurs, looking specifically into the impacts of public policies on that behaviour. Some of the issues addressed by these groups include microeconomic analysis of industrial innovation, the entrepreneurial network seen as a complex system, innovations and the labour market, and innovation and the organization of labour. The research groups work on a wide set of actors: firms, entrepreneurs, the scientific community, academic institutions, public institutions, national and international financing agencies, and people in general, in the latter case through STI public perceptions analyses. In general, these studies look especially at the relationships and interactions between different types of actors of the NSI.

24 Selection of research projects by the innovation groups It would be hard to find any isolated set of specific criteria able to reflect how the research groups select their projects. Rather, selection criteria seem to respond, for each group, to a complex combination of internal and external elements. The relations between the internal and the external elements can lead to virtuous circles for the intellectual growth of the group, though it sometimes can provoke a lock-in in certain topics and approaches, depending on the specific conditions and the characteristics of the groups. Among the internal elements, all groups indicated that their own research interests were a sine qua non condition to select a project or a research line. These interests can derive from individual preferences or they can be the result of a collective process of agenda building. This process is not limited to the group members only but sometimes includes external researchers and contacts, national and international. In the following case, to the extent that national and international topics were incorporated, the collective grew and included more stakeholders in the design of projects: - “RedeSist has always defined its research agenda collectively. Obviously, this is also negotiated with possible funding sources. Nevertheless, the agenda is set by a collective, which from the very beginning included not only the Brazilian part of RedeSist but also other Latin American partners. In parallel, other Brazilian partners became part of the discussion so that our agenda also became more and more national in scope.”

The research projects always aim at accumulating theoretically, given that this contributes to strengthening the group and developing their specific area of knowledge. Within this frame, demand from actors who are external to the group is often welcome: - “We have here an interaction: each research line responds to interests of the researchers and at the same time it responds to interests expressed from the demand side. But, as a rule, we make consultancy work only if the problems are to some extent related to those on which we are working. To give an example, the last work I have done is in the area of health. This was born by an interest of the Pan American Health Organization and by my own interests to enter into the field of innovation in hospital services from a very specific research angle.”

The role played by specific knowledge demands from different actors is, indeed, most relevant among the external elements that shape what the groups ultimately decide to do. On the one hand, the already mentioned interest in exerting influence, from the knowledge produced by the group, on the STI policy definition, implementation or evaluation is an incentive to respond to government demands. On the other, demands from external actors often open roads to new internal directions of research that sometimes induce new external demands, shaping a virtuous spiral. - “We started studying the dynamics of research groups through a direct demand from an OEA sponsored regional project. From this, we went to an internal line on how research was organized. We started working on issues of research and social inclusion, we wrote some papers that went well, and we then started interacting with international partners. Now we work on this issue both at national, regional and international level. The same happened with RedeSist where we started as regional partners providing the view of a small country in relation to NSI and later developed a whole line of reflection on the matter.”

The possibility of actively participating in international academic circuits is another important external element in selecting projects. Finally, the role of financing opportunities is paramount indeed, especially when the group survival depends on external funding. Sometimes, this can have negative consequences, particularly in contexts where few research groups are present, because “following the money” can lead to less variety of topics and theoretical approaches or to a dispersion of research activities to cope with small and diversified projects, each from a different funding agency. In this sense, several researchers expressed the importance of public universities, given that they assure the “ideological

25 freedom" of the research approaches, and represent an opportunity not to follow fashions or select projects for the only reason that there is where money for research is available.

2.2

Researchers’ perception of the use of their research in the PMP

As was said, when asked if they intended or pretended influencing STI policy, all researchers answered positively and usually considered this as a specific objective of their group. In one way or the other, they expressed that the objects of their research was of interest to the policy-making process and therefore wished their findings to be taken into account: - “Essentially, the idea of the group is to generate information and applied research that is useful for decision-taking by government authorities and also at the international level.” - “*I expect an impact+ either through contributing to the construction of development mechanisms or, indirectly, through evaluation work … generating elements that allow the government to improve the functioning of the technological policy instruments. Clearly, one has this pretension.” - “We spend our lives studying the rationality of different STI actors, so we wish to have an influence…”

Therefore, we investigated, on the one hand, how researchers believe they influence the PMP and through which mechanisms, and, on the other, whether they consider their work has effectively had an impact in the PMP. 2.2.1

How do researchers pretend and intend influencing policy-making?

Interviewees were invited to explain how they intended to influence policy design, suggesting them to refer both to direct and indirect ways. In the latter case, we sought to detect the role that mediating people or institutions would play in knowledge transfer or exchange, such as brokers, ‘translators’, or, more generally, knowledge purveyors. Let us say from the start that there were very few references to this kind of intermediaries, except to acknowledge the lack of such figures. Researchers reported different means to influence policy design. The classical publication of research outcomes in books, journals and reports was not disregarded as influential mechanisms, for instance, indirectly, to obtain or consolidate legitimacy as a reference group, or to inform collective actors who can then mark a position on certain topics. Unsurprisingly though, most researchers considered the following means to be particularly valuable: their participation in seminars, forums, round tables, and workshops, where PM are included among the guests; their participation in advising committees or specific commissions related to policy development; and, in a variable degree, their personal discussion of research findings with STI public authorities they are acquainted with. The following interviews’ excerpts illustrate these views: - “Obviously, we expect to have an influence through publications but I wouldn’t say this is the main way. Though it often takes us more time to organize forums and seminars with decision-takers, this allows the exchange and confrontation of opinions in a more proximate way. It can be more important than handing out reports, though of course we also write documents for these meetings.” - “The other way to have an impact *on policy+ is through advising and accompanying institutions and government bodies involved in public policy-making related to the areas, lines and projects of the research group. For example, our national statistics institution was concerned with the measurement issue, so it set up an expert group to help understanding the innovation process. I am advising them on innovation in general and, in particular, on innovation in services. In this type of spaces one has an influence by contributing one’s knowledge and experience”. - “From time to time, people from the ministry call us to know our opinion… Some weeks ago, I had a meeting with the undersecretary, who wanted our group to help them thinking on the logics of policies. It turned out that the topics they mentioned were not of particular interest to our group, but the mere fact that they would call on research groups to help them thinking is an interesting and valuable demand.”

26 - “*I seek to influence+ (i) trough courses, conferences, etc.; (ii) providing advice to government agencies; (iii) constantly following the public debate and seeking to intervene in it; (iv) advising business associations, clusters, etc.

Some interviewees mentioned the importance of the media (radio, press) as a diffusion mechanism of their ideas: - “About once a month I publish something related to STI in newspapers. We also organized a divulgation cycle last year… I don’t mean this is much but it is a lot more than 20 years ago. Anyway, comparing to other fields of social sciences, social studies of S&T are relatively marginal.”

Regarding the media, a few researchers questioned their own capacity as communicators considering that, in addition to being a time consuming task, effectively translating research findings into understandable and meaningful ideas for a non-academic public requires a special ability. Anyhow, it is striking that none of our interviewees referred to holding blogs, wikis or other services derived from the Web 2.0. Impact is not expected in the short term. In fact, interviewees implicitly (but definitively) associated their influence on policy with the conceptual or ‘enlightenment function’ of research (see chapter 1, section 1.1.1). The perception of a ‘percolation’ type of contribution −i.e. the use of new concepts that gradually penetrate in policy networks and alter the language use, shaping the policy discourse− is particularly clear in the following citations: - “Looking at certain concepts that were relatively new in the country, we identified the need to generate a common language in order to have ever more valid interlocutors. This means handling concepts repeatedly in different groups until they are appropriated. We intended this mainly through our active participation in seminars and workshops… We *also+ created mechanisms that allow us to be in contact with decision-takers, with technicians, so as to discuss ideas and concepts and reach this objective of common language, thereby increasing the possibility that our proposals land in a more fertile environment.” - “PM from *different STI agencies+ make an extensive use of our reading material. When I read documents written by them I can see sentences that are ours, I recognize our ideas, without doubt. (…) We repeat concepts over and over again in meetings, seminars, etc. I see how they later incorporate these concepts, I see an evolution, of course not only because of us but I believe we played a role in this. (…) [The problem was that,] when PM attended international meetings, they brought back the framework of thought and different concepts on innovation that have become the standard in international organisations, such as OECD, the European Union, IDB, etc. It was like a rain of concepts falling on them, which they interpreted their own way. Often, their interpretation was mistaken because they didn’t have the appropriate background, but they somehow applied all this and then thought that a NSI was being created. I believe we have been helping them to get a better understanding of these concepts; we do this job of raising consciousness.” - “... It is not the fact that one produces a methodology and the government applies it; rather I believe one generates ideas and contributes to the construction of a [framework of] thought. And this thinking ends up spreading… It is not the work itself, it is no what is written, it is not quantified. Often, one’s participation and impact, or one’s interference with policy design, runs through informal mechanisms… While I’m doing this evaluation of new instruments for technology development, I go to institutions looking for information and I talk to people, and people in one way or the other listen to you, get to know you. They tell you they are doing this or the other and ask you ‘what do you think about it?’ That’s the sort of way, you see, it is not for what is written.” -“… the ways to influence don’t always pass through written things; there are questions very much associated to a text but there is also what is called ‘a state of opinion’.” - “Being able to influence and contribute to policy is an objective, but this might well occur in the medium or long term; it is not immediate.” - “I would say *studies are] a sort of basket that is available for decision-takers, we dialogue with them, we know each other, they know our work to a certain extent. These studies have given the opportunity for this dialogue, we don’t go there talking anything, we have researched this and that. I

27 believe PM do take us into account, moderately. However, policy decision-taking doesn’t depend only on theoretical inputs that a research group generates, rather it depends on many other things…”

A particular way of influencing policy derives from the training of specialized human resources within innovation research groups or centres, or from the participation in these groups of researchers educated in renowned university departments (e.g. the SPRU at the University of Sussex). Intentionally or not, this establishes a sort of ‘reservoir’ of specialists that government agencies might eventually call upon and appoint. - “*At one moment+ the centre, where the national scientific policy was initially conceived, went through a stage clearly linked to policy: the first specialists at the national level graduated at the centre and, later, went to work at the planning ministry. (…) Where I presently work, we also try to have graduates incorporated into the institutions. We try to place them as professionals, to see if they can influence policy.”

The linkage between SSH research and policy-making resulting from people’s movement from academy to the PM sphere and vice versa (i.e., a nexus based on persons, beyond specific research outcomes) was mentioned spontaneously in different countries, though it was specially highlighted in Brazil. A researcher asserted: - “There has been a transit between the academy and the government: from my department, two ministers came out and at one moment we had more than then professors working in the government. (…) [The influence] happens through a symbiosis between government people who are in the academy and academic people who are in the government. It is not strictly due to the academic work but rather to the presence of that the person… People’s experience is assumed to be a decisive factor in the policy configuration, but then also their network of contacts in the academy. When some topics raise doubts, presumably they will activate their networks. I think it is much less probable that a paper published in the Revista Brasileira da Inovação will have a significative impact on policy.”

Similarly, a Brazilian PM expressed: - “I believe SSH have a fundamental role, inclusive because many academics become policy-makers… So, there is an integration between the academy and the government in this debate on innovation and innovation policy... it is more than integration, it is a superposition of persons between two areas. [Would you say this happens because academics take on government positions or rather because there are institutional links between academy and government?] I think both things happen. There are people going from the academy to the government and from the government to the academy; [while] institutionalization occurs more through agreements and the commissioning of studies.”

However, this type of academy-policy nexus is not necessarily viewed as ‘beneficial’ to innovation policy-making. One could tentatively suggest that a distinctive feature lies here in the type of scientific background of PM: hard sciences versus SSH sciences. Indeed, in some countries the fact that STI authorities stem from hard sciences was viewed by researchers as leading to an ‘imbalance’ between S&T and innovation policies. Different factors might intertwine to produce such an effect. On the one hand, hard sciences researchers often constitute a strong community, more prone to collectively champion their interests than SSH researchers, especially in the STI field. Their voice is then particularly convincing when their own peers sit on the other side of the counter. On the other hand, in most Latin American countries the design of effective and articulated instruments to promote innovation in the business sector and in society is still in its infancy. So, PM with a scientific background might perceive lesser risks (and feel more at ease) when designing S&T policies and instruments than in the case of innovation. Finally, PM from hard sciences might not be naturally driven to exploit all that social sciences has to offer, while at the same time the supply of SSH tends to be scattered and unarticulated in most countries, resting to their strength as will be analysed farther on. Researchers also referred implicitly to the instrumental function of their research, i.e., research providing empirical evidence that helps solving a policy problem. This happens when research groups respond to specific demands from the public administration (research contracts/agreements

28 or consultancies). Practically all innovation research groups respond, to a higher or lesser extent, to requests issued from government agencies, besides their main research functions (see section 2.1). Though the evaluation of STI policies, instruments, and programs is reportedly a weak area of most Latin American governments, some interviewees had been commissioned to this specific end and considered this to be a particularly important use of their research capacities. In this case, the diffusion of research outcomes is sometimes viewed as a task that lies outside researchers’ responsibility: - “... *N+ormally, I dont’ worry then about diffusion because these studies are commissioned and are of an institutional character. The contracting institution is thus the one in charge of the diffusion of these studies, ideas, proposals, or conclusions.”

Few researchers referred to what was defined in section 1.1.1 as the ‘symbolic, strategic or legitimative use’ of research, or what Weiss (1977) calls the ‘political model’, i.e., when research becomes ammunition for PM. One interviewee expressed: - “In the case of some researchers… they are commissioned to backup plans or decisions. *Would you say that there is a political criterion?]. Absolutely, the ones commissioned were the same people who had expressly declared there support to the government.”

2.2.2

Can researchers pinpoint policy impacts of specific studies?

Interviewees were asked to identify one or two studies of their group that they considered had become a reference in the STI arena, either locally or internationally. They were then requested to indicate if these studies had been taken into account specifically in the policy sphere. Unsurprisingly, the answer was not always affirmative and sometimes researchers doubted about the real impact of the study by itself. At times, the answer was positive but lacked concreteness on the impact. Table 2.1 provides a series of research outcomes for which researchers were definite about the impact at the policy level. In each case, the exact bibliographic reference is indicated in a footnote. Accordant with the previous section, reported impacts are a mix of instrumental and conceptual use of knowledge production. A couple of examples refer to non-linear impacts, through intermediaries such as an advocacy group or the business sector. Also to be noticed is the fact that several of these influential studies were collective publications, either of innovation groups or ad hoc groups. Table 2.1 – Examples of uptake of research outcomes at the policy level in selected Latin American countries TYPE OF WORK

INSTITUTION

IMPACT IN NATIONAL PUBLIC POLICY

Study on the emigration of human capital16, 2002

Centro REDES (Argentina)

Gave rise to a policy instrument (goal of 1.500 doctoral fellowships per annum)

Design of a program for SME technological counselling17, 2004

Instituto de Industria, UNGS (Argentina)

Translated into an instrument after adaptation by the S&T Secretary.

Concepts and studies on productive tissue and innovation (Tramas productivas e innovación)18

Instituto de Industria, Universidad Nacional General Sarmiento (UNGS) (Argentina)

Taken up by the Ministry of Labour. Motivated more studies with these concepts from a labour perspective.

16

Albornoz M, Luchilo L, Arber G, Barrere R y Raffo J (2002), El talento que se pierde. Aproximación al estudio de la emigración de profesionales,investigadores y tecnólogos argentinos, documento de trabajo nº 4, Buenos Aires, Centro de Estudios sobre Ciencia, Desarrollo y Educación Superior (REDES). 17 Yoguel G, Neuman M, Gatto F, Malet Quintar Braidot N y Nicolini J (1997), “Programa de mejoramiento de las capacidades tecnológicas de las Pymes”, en Plan Plurianual de Ciencia y Tecnología 1998-2000, Documento Nro 1, FONTAR, SEPCYT.

29

Project/study on “Unapplied Applicable Knowledge“ 19, 2004

Universidad Nacional de Quilmes (Argentina)

No impact at government level but was taken up by an advocacy group20.

Study on institutional added value in human capital formation in clinical research21, 2008

Instituto de Economía, Universidad de Rosario (Argentina)

Conceptual impact at the STI governing body, COLCIENCIAS.

Project/study on globalization and local innovation22, 1999

RedeSist, Instituto de Economia, Univ. Federal do Rio de Janeiro (Brasil)

Helped focusing innovation policies on local, micro systems of innovation

Study on the promotion of technological innovation in SME23, 2004

APEI with the advice of researchers of Univ. de Sao Paulo (Brazil).

Indirectly, through a broad diffusion and acknowledgement in the business sector.

Impact studies on STI programs/ instruments

INTELIS (Chile)

Taken into account to tune up different programs/instruments.

Document with recommendations on governability issues of the NSI

INTELIS with a World Bank working team (Chile)

Inclusion of several recommendations in Innovation Council documents to orient the NSI and policy instruments.

Three books on the outcomes of the first innovation survey (1996)

OCYT (Colombia)

No direct impact when published but contributed to diffusing new concepts.

Production and analysis of national STI indicators (2007)

Collective work of a special commission created by the S&T Ministry (Costa Rica)

Contributed to the shift from a R&D approach to a RD&I approach.

Diagnosis of STI policy in Mexico 24

Commissioned by the ‘Foro de Innovación’ to a technical group (Mexico)

Important input in the broad discussion of the official STI program

Project/study ‘CIENTIS’ (STI diagnose and policy proposal), 2003

Collective work including about 300 researchers, entrepreneurs, university authorities and external specialists (Uruguay)

Constituted one of the five axes of the program of the then opposition political party (Frente Amplio), which took up government in 2005.

18

Several publications. A recent one: Yoguel G. (2007) “Tramas productivas y generación de ventajas competitivas: un abordaje metodológico para pasar de la firma individual a la red”, en Novick M y Palomino H (ed.), Estructura productiva y empleo, Buenos Aires: Miño y Dávila, Ministerio de Trabajo. 19 Proyecto: “Construcción social de la utilidad de los conocimientos científicos y tecnológicos en contextos periféricos. Una indagación sobre el fenómeno de producción de Conocimiento Aplicable No Aplicado (CANA)”. PICT, Agencia Nacional de Promoción de la Ciencia y la Tecnología (FONCyT), 2004-2006. Kreimer P y Thomas H (2004), Producción y uso social de conocimientos. Estudios de sociología de la ciencia y la tecnología en América Latina. Buenos Aires: Editorial UNQ. 20 Grupo de Gestión de Políticas de Estado en Ciencia y Tecnología. 21 Jaramillo H, Latorre C, Albán MC, Lopera C (2008). “El Hospital como organización de conocimiento y espacio de investigación y formación. Los recursos humanos en salud y su tránsito a comunidades científicas: el caso de la investigación clínica en Colombia”, Centro Editorial Rosarista, colección textos de economía, Facultad de Economía, Bogota. 22 Cassiolato J y Lastres H (eds.) (1999), Globalización e Innovación Localizada. Experiencias de sistemas locales en el Mercosur, IBICT, Brasilia. 23 APEI (2004), Como alavancar a inovação tecnologica nas empresas, Associação Nacional de Pesquisa e Desenvolvimento das Empresas Inovadoras (APEI), Sao Paulo. 24 Dutrénit G, Capdevielle M, Casas R, Puchet M, Unger K, Vera-Cruz AO et al (2006), Diagnóstico de la Política Científica, Tecnológica y de Fomento a la Innovación en México (2000-2006), FCCT: México. Dutrénit G, Capdevielle M, Corona Alcantar JM, Puchet Anyul M, Santiago F, Vera-Cruz AO (2008), The Mexican National System of Innovation: Structures, Policies, Performance and Challenges, Background Report to the OECD Country Review of Mexico’s NSI, Technical Report, Conacyt.

30

Study on the situation, prospective and policies in STI, 200425

Collective work of eight researchers of the Universidad de la República (UDELAR) (Uruguay)

Input taken into account by the executive and legislative powers in the design of STI policies.

Study/proposal for a knowledgebased development strategy26, UNDP Human Development Report 2005

Study commissioned by UNDP to Instituto de Economía, UDELAR (Uruguay)

Contributed to the uptake of the concept of productive transformation through transversal sectors, at the policy level

Concepts related to the need of linking innovation and social policies (social inclusion), 2005 onward

Comisión Sectorial de Investigación Científica, UDELAR (Uruguay)

Uptake of the basic concepts in the STI national strategic plan and in the design of instruments for social inclusion through innovation

Study on the national electronics sector, 1992

Centro de Informaciones y Estudios del Uruguay (CIESU) (Uruguay)

Contributed to declaring of national interest (fiscal exemptions) the electronics sector in Uruguay

Manual de Bogotá27 (conceptual and methodological adaptations of the Oslo Manual to developing countries) , 2001

Collective work with the participation of researchers of several LA countries

Uptake of concepts in national industrial innovation surveys in different LA countries

STI national and regional indicators, and comparative view of the STI institutional frames in Latin American (periodic publication) 28

Centro REDES, in the frame of RICYT (Ibero-American and Inter-American Network for S&T Indicators)

Basic STI indicators (by country and comparative) used as a reference tool in most STI diagnoses and other policy related studies.

Source: Interviews to members of SSH research groups on innovation in LA, 2009-2010.

2.3

Researchers’ view of the obstacles to a closer research-policy nexus

Researchers were requested to identify the obstacles they perceived in research-policy linking. After revising all answers, we considered the applicability of the ODI-RAPID model (Box 1.4, chapter 1), which groups the key variables of research use in the PMP into four categories: context, evidence, linkage and external influences. We finally opted for a similar classification, with some variations according to the problems effectively identified in the interviews (Graph 2.1). Obviously, the systemic nature of the question at stake means that any classification is, to some extent, arbitrary: some difficulties belong to more than one category. It is also worth recalling that the list of problems is not necessarily exhaustive, neither does it follow a hierarchical order. Rather, it gives account of the spontaneous opinions expressed by renowned researchers, either to the specific question or through previous comments during the interview. Table 2.2 intents to tag the problems, which are hereafter elucidated by means of interviews excerpts

25

Bértola L (Coord.), Bianchi C, Darscht P, Davyt A, Pittaluga L, Reig N, Román C, Snoeck M, Willebald H (2006), "Ciencia, tecnología e innovación en Uruguay: diagnóstico, prospectiva y política", Serie de Notas de referencia, RE1-RN-05-001, BID, Washington.. 26 PNUD (2005), Informe de Desarrollo Humano de Uruguay 2005. ‘El Uruguay hacia una estrategia de desarrollo basada en el conocimiento’, PNUD, Montevideo. 27 Jaramillo H, Lugones G, Salazar M (2001), Normalización de Indicadores de Innovación Tecnológica en América Latina y el Caribe. Manual de Bogotá, RICYT-OEA-CYTED, COLCIENCIAS/OCYT. 28 REDES (2009), El estado de la ciencia. Principales indicadores de ciencia y tecnología,RICYT.

31 Graph 2.1 – Types of difficulties hindering the SSH research and innovation policy nexus according to researchers

Evidence

Nature of PolicyMaking Process

(limitations)

(complexity)

Links (mismatches)

Intertwined obstacles to "SSH research and innovation policy" nexus

Governability and governance (difficulties)

External factors (influence)

Source: Interviews to members of SSH research groups on innovation in LA, 2009-2010

Table 2.2 – Checklist of obstacles to a stronger research-policy nexus in the field of innovation (as reported by SSH researchers for their respective country) On the research side (evidence) Frequent lack of convincing power of SSH research results. Short-sighted researchers' view of the innovation process, limiting their perception of (institutional) restrictions of policy-making and, therefore, their findings validity. Divorce between economic, political and public policy view of innovation: an economic-based frame of thought prevails, with limited interactions with research on public policy issues. Among economists, the equilibrium model view of the macro economy prevails over the institutional and micro approach (à la Freeman and Lundvall). Tough dialogue between social scientists and lawyers to solve coordination problems. Innovation researchers are predominantly ‘industrialists’. Innovation in more dynamic sectors has been neglected. A disciplinary research mode thus still prevails over a problem-based and multidisciplinary approach. Lack of human capital specialised and trained in innovation topics. On the linking Mismatches between research supply and demand Knowledge supply from SSH (indicators, base studies, etc.) is subutilized in the decision-making system. Research is auto-referential: researchers are reluctant to take advice on knowledge needs of other actors. Two communities problems Different languages (jargon) entail the need to decodifying on both sides. Timing and evidence requirements of PM and researchers are distinct. The academic evaluation and incentives system is contrary to researchers’ involvement outside the academy. Historical and/or political factors exacerbate the gap. On the nature of the policy-making process The many stages following STI policy design to reach decision-taking are outside the reach of researchers. Beyond theory, when it comes to define policy we are all actors with our own interests, inertias, tramps, etc. Ideological, strategic, tactical, circumstantial and personal factors interplay in policy-making and priority setting. PM want fast and simple evidence, and ambitious STI plans to leave their imprint. Little interest in learning from

32 previous strategies and instruments. Agenda problems: a change of government or unforeseen events may suddenly affect STI priority in the policy agenda. Research findings do not easily permeate when adverse to preconceived ideas, subjacent to some PM actions. On governability and governance The 'principal and agent' relation affect STI policy and the crucial articulation between public policies and instruments. Lack of consultation tradition of PM, and of citizens' participation in STI public issues. Low empowerment of STI ministry. PM demands are discretional; they are often addressed to privileged groups. Dialogue and meetings taking place between different actors are inefficient in terms of knowledge exchange. Difficulty of collectively building an articulated policy, encompassing other social actors than PM and researchers. Lack of articulation of macro, sectorial and STI policies. Ensuing inconsistency of instruments. Institutional restrictions. Lack of a specialised bureaucracy in STI, trained to taking into account research findings as inputs for policy design. Public agencies do not exchange information on the findings of the projects they finance, to improve policy design. Public policy decisions are frequently taken without information and knowledge. Limited and/or discretional diffusion of primary data obtained by public entities (surveys). Lack of development strategies whose long-term objectives require a focus on STI (e.g., structural change). External influences Latin American mimetic: solutions adopted in the North are replicated as if problems were identical in the South. International research networks influence the setting of local research agendas. Neo-liberal times left behind a remnant of supply based policy. Source: Based on 39 interviews to members of SSH research groups on innovation in LA, 2009-2010.

2.3.1

Limitations on the research side

Only one informant, from a developed country, referred critically to the nature of the research outcomes: "… *t+he analysis provided by social sciences research often is not fully convincing".

More subtly, an interviewee from LA pointed to the importance of consolidating the position of SSH researchers and of maturing research strategies in the field innovation: - "… a first point is that we should consolidate our position. We have been progressing but it is all still modest, it’s not that anybody would breathe without us! [We should] professionalise our own work, progress in our researches, offer more findings, getting more mature work strategies… The first obstacle is that we need to continue growing wiser."

Undoubtedly, establishing credibility and legitimacy through excellence in production, as a necessary condition for being taken seriously in any arena, and a fortiori in the political one. When research findings are inconclusive, limited in scope, out-of-date, or contradicted by other studies, no clear message is being conveyed to PM. The existence of a critical mass of researchers and research groups that are visible as valid interlocutors in different innovation topics is another needed condition to eventually attend PM demands and institutionalize the linkage: - "… the problem is the lack of a critical mass of people thinking on S&T policy topics that are hosted in different visible units where they can be consulted or convoked. Right now this happens perhaps more on a personal than an institutional level. Mi diagnosis is that this is due to the lack of persons who are studying these topics in the country." - “There is no critical mass, in the first place. If we look at the European countries, the number of people working on these issues is incomparable with the Latin American case.”

Other obstacles that were highlighted referred to the still very unidisciplinary character of research on innovation. Several comments pointed to the prevalence of an economic approach that leaves

33 relatively little space to political sciences, as well as to barriers in the dialogue taking place between social scientists of distinct disciplines. The following quotations are illustrative in this regard, besides insinuating some other problems that are treated in other categories: - "One type of difficulty is that a majority of academicians studying innovation have a short-sighted vision of this phenomenon and, therefore, don’t perceive the restrictions on the PM side. Consequently, researchers’ suggestions tend to be less effective." - "The economy has been considerably monopolized as social science by macro economists with an equilibrium conception of society’s way of functioning. This continues to be that way today in spite of the belief that things have changed in recent years. The most we have progressed in economics is recognizing illustrated economists as Krugman, Stiglitz or Rodrik, but the spectrum of the institutional and the micro –namely in the sense of Freeman or Lundvall− is still little accepted in the mainstream economic circles. It continues to be a marginal debate, sometimes discarded because of the suspicion that it either hosts protectionist feelings or ideas that are foreign to a market-based reading of the social process.” - "We are used to formulate things in terms of an analytical framework stemming from economics, the ‘theory of innovation’, and … we don’t take into account a neighbouring branch, the one that studies public policies. (…) One should be capable of incorporating in our proposals many of the obstacles, limits and conditioning established by the policy-making process. We have researchers who have studied this empirically in the country; they know the obstacles and limits to policy outreach. Mi impression is that we are in a condition to accomplish this integration [between research groups on innovation and on political sciences+.” - "We have a problem when dialoguing with lawyers. The legal structure is pyramidal by definition; it is organized on a constitutional base, which establishes the rules of the subordinated bodies. But, when one has to think a system of rules and a regulatory frame for an innovation related issue, one is obliged to think transversally. Inserting a new pyramid, which will organize another bureaucracy, might prove very efficient to a certain effect but most probably it will compete with an existing pyramid aside and maybe several of them, so that it does not solve the coordination problems, which are the relevant ones here. Thus, it is very complex to transit from discussion to the design of proposals that are not of the style: ‘Let’s create another council, on top of the existing councils, by mean of a constitutional arrangement’. For example, Mexico is now strongly pushing ahead the creation of a ‘super’ ministry that would group science, technology, higher education and innovation. This will be an amalgam of pieces of bureaucracy to make a larger bureaucracy but the coordination problems, for instance with the university system or with the federative entities, will not be solved by establishing a new, higher, and larger pyramid, no? … The easy way is always ‘let’s draft a larger law’… The main problem is not coordination inside the public administration (which obviously is also a great problem) but coordination between the actors involved in innovation, because they are located in different organizations and respond to distinct systems of rules. I mean, how do business people articulate with researchers, with the civil servants in charge of policy-making, etc. We don’t have any proposal for this but, for sure, you cannot issue a law forcing all these actors to behave this way or the other… We need to study their own systems of formal rules and see how to remove all that is of no use anymore or introduce modifications with a fine tuning exercise. I believe this is a topic in which we, researchers, have a very serious pending issue.” - "Throughout the last decade, the level of innovation in the agricultural sector of Argentina has been tremendous. It should be the object of very strong theoretical studies. In soybean production innovation has been related to biotechnology, climate, soils, nutrient, trade networks, agricultural equipment, etc. Much of the agricultural machinery is produced in the country and has become a potent industrial sector internationally. All these innovative, competitive, and far-reaching changes have been very much disregarded by researchers. Besides the existing divorce in the country between the city and the countryside, all theoreticians of innovation are specialized in industry.

Some researchers recognised a tendency to define their agenda in an auto-referential way: - "… we don't ask anyone what should be researched; we tend to believe that we have the capacity to define the research agenda in an auto-referential way."

34 Others referred to the separation between the hard and social sciences spheres, like in this rather radical comment: - “Social scientists don’t participate in S&T policy, that’s the territory of hard sciences; it is acknowledged as hard science territory by the social scientists. The latter participate in health, habitat, etc., policy-making, which are their object of study… This occurs at the national and the university level. For example, I wrote an article a long time ago … where I said that the university research policy was built in a totally atomized way: each group decided what it was going to do and the social scientists accepted breadcrumbles because the bulk of the funds was assigned to hard sciences and they didn’t want to interfere.”

Finally, connectivity between innovation research groups is weak, within countries and within the region. Researchers report a fluid relation with most of their peers, with whom they meet mainly in seminars or similar events. However, with a few exceptions, this does not entail a working relationship between groups or people, as the following comments demonstrate: - “In general, the linkage is not so strong… the groups don’t interact much with each other. It occurs, for example, that [national] researchers meet more at some of the seminars abroad than here, and this reflects the disconnection… We maintain good relations, we read papers that others write and from time to time we participate in seminars with them. But we don’t work together, there is no joint work. For example, there is no such thing as a research program between people doing similar things. I believe this has to do with several factors: on the one hand, there is no habit of connectivity and, on the other, there is some institutional jealously due to the fact that some groups finance their activities with studies based on primary information that others have no access to.” - “*Connectivity+ exists, it is more or less intense but less important than what should be. In a certain way, for several reasons, more sociological than of any other order, there is a certain ‘balkanisation’, I mean, a greater rivalry than desirable. So, there is acquaintance but it is not deep enough or sufficing to establish a very good dynamic. The research groups have more relations with groups from abroad than from inside the country.” - “Two years ago I thought about creating a monthly seminar on innovation topics. I looked for people and institutions and invited them but got no answers. Also, we have an annual meeting of economists but there does not seem to exist any interest in these topics.”

2.3.2

Mismatches between research supply and demand

As was highlighted in chapter 1, there is a tendency in Northern countries toward 'evidenceinformed policy dialogue and processes'. Consequently, S&T information acquires increasing importance in the PMP, for example, in the fields of health, natural resources management, agriculture, climatic change, food security and biotechnology (Jones, Jones and Walsh, 2008). In spite of the multiple barriers for research uptake in these countries, and the pros and cons of evidence-based policies, it could be said somewhat roughly that hard sciences play an instrumental role (hard facts) while SSH contribute to the understanding of social phenomena, among others the complexities of PMP and the role of innovation in the economy. There are institutionalised mechanisms that favour or facilitate the research-policy nexus, such as international debates in the frame of the EU and diverse international organisations, where questions on the demand side meet with those of the supply side. In LA, the challenge implied in strengthening the research-policy nexus is greater for different reasons, including those that are reported in this chapter. But, a very basic issue is the mere fact that PM themselves do not have a clear idea of the knowledge needs of different local actors, in the context of underdevelopment. Common sense would indicate that this makes up a trigger for an articulated and organised demand toward SSH, especially to conceive mechanisms to detect needs and transform them in explicit knowledge demands. However, our interviews indicate that this has not happened, at least in the field of innovation; demands addressed by PM to researchers tend to be punctual and oriented toward filling particular knowledge gaps.

35 In brief, SSH research supply and demand run in parallel paths with relatively few crossing points. While research supply is relatively limited in LA (according to the number of innovation research groups that we identified) and inarticulate, on the whole research demand is imprecise, indefinite, unclear. The following comments illustrate this: “… According to one of our studies, researchers find that a barrier to their contribution to national development consists in the lack of early information on the type of problems that exist, i.e., not on the priority lines but rather on the type of problems they could work on. Therefore, one of the tasks at the policy level should be to build a ‘supermarket of problems’… There is a distance between the capacities to solve problems and the institutional conditions that are needed to put these capacities to work on the problems… There are no broad institutional spaces where to discuss these things, beyond their presentation in seminars after which everyone goes back home and nothing happens.” - "… STI decision-takers show a low level of demand for knowledge generated in the SSH field; they think they know it all... I believe there is little demand for indicators, base studies, etc. All this is scarcely used in the decision-taking process. They are typical products of a supply oriented system: information that was not demanded in the decision-making system". - "… policy-makers are unaware of the studies that could illustrate their own work" - "… policy-makers don't consult much with researchers about the things they do".

Naturally, several interviewees pointed to the classical academy-policymakers communication and articulation difficulties, the same that were put forward in the "two communities models" (Box 1.4, chapter 1). Differences in language, timing, interests and, especially the incentive systems were explained in the following terms: - "... there is no good translation of research outcomes that would help PM to transform this in concrete lines to the ground. (…) Neither PM know how to decodify what researchers do, nor researchers have the least interest in presenting their questions in terms that could be much more useful for policy design. Researchers are more interested in writing a paper and publishing it, and, in fact, their qualification depends on the quantity of papers they publish. Nobody is evaluating in the scientific system, for instance, to what extent some idea of some researcher can be a key to solve a policy question, to solve people’s problems, to eradicate endemic illnesses, or to think about instruments that were not implemented before. (…) The incentive regime is not conceived to value the interconnectivity of a researcher with the public policy system in a way that it is useful for his academic career. There is no reason to think that the only important thing is publishing papers: to solve people’s problems seems to me important as well.” - "On the one hand, there is a problem of interpretation of research results, of what could be really their impact in public policy. Often the academician does not succeed –due to the very lack of knowledge of what the public dynamics are– in presenting research outcomes in terms of how to influence public policy.” - “… researchers do not have many incentives to be in the public discussion, basically because what is valued is his teaching task and his publishing.” - “The main obstacle in my view is language and also times.” - "Though it may sound silly, the time mismatch between the ones and the others ends up being very serious. PM have limited time spans to produce outcomes and draft policy documents... Even when existing studies include important issues, the policy-making life makes it difficult to think about the latter and on how to incorporate these in their proposals. This is a fact in our country.

This last comment reminds Cable (2003) stand on the five S's that limit evidence-based decisionmaking, according to his own experience as a Member of Parliament in England: speed (PM have to make decisions fast); superficiality (PM cover a wide brief); spin (public perception is very influential in political decisions); secrecy (many policy discussions have to be held in secret); and scientific ignorance (few PM are scientists and understand, for example, the concept of testing a hypothesis).

36 2.3.3

The nature of the policy-making process

Interviewees on the research side are of course well aware that political processes are complex, multifactorial and non-linear (see chapter 1). This awareness is subjacent to their view of SSH influencing indirectly and conceptually the PMP rather than instrumentally, as was previously illustrated. In the present category, we have tried to reflect some powerlessness feeling of researchers in front of the very nature of the PMP. -"I think the main difficulty lies in the very nature of how political decisions are made. One thing is designing: researchers can do this. However moving from design to final approval entails the entire political lobby, the power struggle. We, as researchers, do not take part in this process, neither is it our responsibility to do so. … We are convinced of the need to raise public awareness by promoting public debates and forums, otherwise projects and proposals wouldn't even reach a first round. But we don't know what can happen in the next stages, where the power game is what counts". - "Obviously, there are power structures in the country associated to individuals and political parties, which in contrast to earlier times now include more people acquainted with the STI field. Those are the people who, according to an interrelated set of ideological, strategic, tactic, circumstantial and other factors, position themselves in a given way in regard to policy, partly because of what they think, and partly because they live in a world of incentives and protection that influence them in doing or not doing certain things at given moments." - "There are theoretical issues but, when we are going to design a policy, we are all actors, with interests, inertias, traps." - "Many PM in the country have an idea of how innovation should be supported and they push this idea ahead, often based on perceptions and very little on hard evidence. I think there is a question of fundaments here, particularly in this field… The alternative discussion of somebody who, showing his findings, says 'look, it could be rather this or the other way' is difficult because people come with preconceived ideas. It is difficult to permeate." - "Politicians not only take decisions fast, based on information and data that are very simple −complexity scares them−, they also wish to implement plans that are very innovative in descriptive terms (to leave their seal) and ambitious (pretending to solve historical problems in four years) so as to cause the highest media impact. They don't evaluate what was done previously: it doesn't interest them since most probably it was done by others (from the same or another party, it doesn't matter). They don't care if it was good or bad, and therefore don't evaluate." - “People don’t last long in their functions … each minister comes with a new idea and comes with his own people.” - "The Innovation Council has operated as a sort of 'accumulator' of information and knowledge, actor's discussions, international organizations’ recommendations… In principle, a good deal of the [evidence-policy] problem has been solved. However, with the present change of government, we don't know if the Innovation Council will be empowered, if the Finance Minister will be interested in listening to policy recommendations in the field [of innovation]. The country has undergone an earthquake and STI issues fell drastically in the government priority list."

2.3.4

Governability and governance issues

Interviewees from several countries seem to consider that research uptake is indirectly limited by the ill functioning of the steering capacity of the state or by a weak performance of strategic actors in the socio-political system. In other words, different issues of governance and governability are at stake, as evidenced by the following comments. Disarticulation of policies, power imbalances, and inefficiencies at the government level: - "A central problem to be solved lies in the governability of the STI system… When the country defines a policy, let say for the science sector, this does not ensure that some policy spaces don’t go against it; some ministries take different policy decisions… The theory of the ‘principal and agent relation’ applies here. At the highest state level there is a policy intention, but within the ministries

37 each agent captures the public policy according to its particular interest, its immediate and short-term interest. The public policy is being captured within the state itself but also within sectors that don't belong to the state." - “It shouldn’t happen that a policy contradicts the other, this is chaotic. It has become fashionable to talk about policy mix: a package of policies that are adjusted through time, trying to harmonise them to reach the fixed goal. This does not happen here. Somebody in a ministry happens to think about a certain policy and somebody in another ministry, something else; so that it finally turns out that export policies have nothing to do with employment policies, etc.” - "There is a lack of articulation between policies at the macro/sectorial levels and innovation policies, … of articulation between different institutional sectors and inside each sector, … and of policy and instruments consistency… For example, for the industrial sector a trade-off has to be found between subsidies stemming from the economic policy (incentives, exchange rate, protection) and subsidies from innovation policy (risk and uncertainty)." - "… Let's take for example the case of the agency that heads both the institution that funds R&D projects and the institution that funds innovation projects at the firm level… Our group relates with both of them, and what we observe is that their linkage is weak in terms of research findings from (applied) R&D that could be used to think about innovation policies. Maybe this has changed very recently but just imagine: if this happens at the agency level, what linkage can we expect between universities and technology centres to discuss policies." - "What we need is a government giving impulse to STI. The problem is that our latest STI ministers haven't exactly been politically sharp-toothed. They came from the academic sector so that they probably understood quite well what was to be done, but they didn't have enough strength in front of the finance, industry or economy ministers." - “I think that STI policies have been limited due to their low valorisation in the dominant conceptions of the political economy of the country... ” - "I believe a main obstacle is the lack of a trained bureaucracy in STI steering and policy-making. We don't have a staff of government officials ready to implement long term policies. Those are the ones who should know the existing academic and technical work on STI and propose different alternatives to politicians. There are some people at that level, but not well trained". This problem was also raised by a research group from the North: "Politicians don't have at their disposal strong and well related study departments, where people would know what each research group in the country is able to do so as to request support in each case accordingly. Therefore, they end up commissioning studies to the people they are closed to or to their friends. At the same time, there is a series of lobbying entities and networks trying to ensure that 'their things' are present in all initiatives (and so it happens, of course)."

Weakness of rules and norms determining the participation and interactions of social actors: - "There is no consultation tradition in public organizations, and citizens participation in government affairs is uncommon. Therefore, only groups with direct and privileged relations manage to have their discourse and research findings taken up at the decision level." - "I am tempted to say that the relevant question is not about the linkage between PM and researchers. … *P+resent STI PM used to be high-level researchers themselves… I insist on the systemic issue: how to encompass other actors to build collectively a joint and articulated policy, i.e., dynamic interrelations that generate actions toward innovation. Sitting them at a common table… In order to know the innovation demands, it is necessary to look for them outside the academic knowledge generation sphere." - "I believe that the distance [between research and policy] rises from the fact that the meetings and dialogue spaces are organized through mechanisms that are essentially inefficient in terms of information transmission." - “Policy instruments are not very much discussed; it is a rather closed discussion inside public organizations. I think this is a very important deficiency… Before launching an instrument, the PM should have some preview of how those instruments will be evaluated, what their expected outcome

38 is, which objective group is aimed, what would be an appropriate control group, etc… Often, the instrument objectives seem plausible but there have been important deficiencies at the design level."

Uneven access of researchers to basic information: - "[Primary] data are not public goods in this country. I think this is a key issue. In Europe, the technology surveys are available to anybody; we don't have access to microdata and we work on many issues that would require them. For example, there is a group working on appropriability and it would benefit from the microdata, of course without the firms’ names. We also work on connectivity questions and we have to do it with our own surveys. But other groups do access the data. I am in favour of public goods: we all should have access to this information, it would improve the quality of the studies and would make the funds assignment process more democratic; otherwise, when there is a call for a study, only the ones who have access to the data can do it." - "Innovation surveys have been beneficial to a certain point, and we must promote that public institutions provide more data that are available; this would increase the linkage between the academy and the PM… We don't have a good access to [microdata from] surveys; you need to know the people who own them."

Regarding these difficulties to accessing disaggregated data from innovation surveys, this has been highlighted as the most often cited obstacle to the use of these surveys in a recent regional study on inputs considered by decision-takers for policy design, implementation and follow-up (Baptista et al, 2009: 24). The study covered Argentina, Chile, Colombia and Uruguay, and in the first three countries this obstacle was reported as a critical factor because it hinders research on topics that are important for decision-taking. Some interviewees also declared that personal relations were often needed to get the data, as a substitution for institutional collaboration mechanisms. 2.3.5

Influence of Northern frameworks on STI policy and research

The influence of Northern frameworks of thought on STI policy and agenda setting in the South (e.g., through international organisations) may affect the research-policy link in the sense that policy priorities determine the type of studies that PM commission to research centres, consultants, etc. On the other hand, these frameworks, together with vested economic interests, might also directly orient research in the South. The most radical position was presented by one of our interviewees in a public debate: "Our research agenda is defined in the central countries, which define their agenda based on their realities. What we do is adopting this agenda … as the frontier science that has to be imitated or emulated. Therefore it is disjointed and not focused on our own reality" (Dagnino, 2008). Another two of our interviewees expressed: - "When a minister gets into office and says 'the priorities are three: nano, bio and TIC', he isn't really fixing priorities, he's just copying the international agenda." - "We do what others have done as if our problems were the same. Francisco Suarez, who tended to use a formal language, used to say: in developed countries a real need emerges in time 1, and in time 2 a solution to this need appears, which can be a new research line, a new profession, or whatever. In the developing, ‘mimetic’ country, in time 1 a replica of the developed country appears, and in time 2, ‘well, let's see what use we make of it’… The problem is not so much that most financing come from international organizations (IDB, World Bank, etc.); it is not so much tied to given lines. The influence is rather on the design of [similar] instruments: all countries have their STI agency, their fund for this and their fund for that. In this country, even with strictly national financing, most probably biotechnology would be a priority; you can’t really attribute this to IDB or another international organization. Would not any self-respecting country create a nanotechnology program, albeit dwarf? I mean, there are issues that are part of the international agenda. The mimetic feature goes beyond what an outsider can impose; underdevelopment itself is more to be blamed for mimetic than imposition by others. "

This brings to mind Hirschman’s (1975: 392) early interpretation of this phenomenon: “I suggested that understanding of a problem and motivation to attack it are two necessary inputs into policymaking and problem-solving, but that the timing of these two ingredients could be significantly out

39 of phase: understanding can pace motivation (as is assumed by Marx and Lindblom), but in other situations motivation to solve a problem may arise in advance of adequate understanding. The latter situation, I argued, is characteristic of Latin American countries to the extent that they import ‘solutions’ from the outside in the form of the most up-to-date central banking legislation, economic planning agencies, or common market schemes. This typically ‘dependent’ behaviour results, of course, in frustration precisely because these institutions are often established without the necessary minimal understanding of the problems they are set up to resolve” (Ibid: 392) 29. Also addressing the question of external influences are the following comments: - "Today, in spite of the generalized evidence that the times of 'state is bad' are over, in practice we still have pieces of bureaucracy saying that the state must be minimalist. In the STI field, this translates into an essentially supply-based policy, sometimes dressed with a look of modernity: 'let's offer much science, much technology to see if this diffuses through the economy and society'." - "… the local research agendas are influenced by the participation of *the country+ in international research networks… They are marked by the large networks and the interests of the European Union, and to a lower extent of some US centres. Maybe this is less so in SSH than in hard sciences."

2.4

Policy-makers’ view and use of SSH research on innovation

Some of our questions to PM concerning SSH research were answered quite in line with what might be expected. For example, practically all interviewees were able to identify several groups, centres and/or researchers working in the field of innovation in their country. When asked if they remembered a couple of SSH studies that had been taken into account in the political sphere, they all cited some studies, with a varying degree of precision and coverage. 30 However, in spite of their acquaintance of existing studies, the imagery of local SSH research was far from being some kind of directory were groups, researchers, and research lines would be clearly linked. In other words, they gave the impression that, on the whole, if they had to consult on a SSH aspect of an innovation issue, they would have to search for information on who exactly in the country is working on this topic, probably through one or the other researcher they particularly trust. In this section we analyse two features of PM’ view on research. First, their expectation from research, i.e., which aspects of STI do PM believe social sciences should focus on. Second, we use the answers of PM and researchers to a question of the survey on STI policy inputs, so as to situate the use of research in relation to other inputs, and contrasting PM and researchers’ view. 2.4.1

Role assigned by PM to SSH research in policy-making

The role PM assigned to SSH was associated with filling knowledge gaps in a very broad range of STI policy matters, especially related to the social nature of innovation processes: - “Presently, many of the actions that are carried out and need to be implemented require, even when the technical proposal is well determined, social sciences work for their implementation. Indeed, society is a very important element to consider. It is not easy to take certain measures without 29

This has not been the privilege of Latin America though, as Hirschman asserted with humor: “The motivation-outruns-understanding style became in fact dominant in the sixties in the United States and unhappy experience with it has then led to the oft-heard complaint: ‘If we can put a man on the moon, why can't we solve the problems of the ghetto? (Nelson, 1974)’. Policymaking on poverty, pollution, cancer, and lately, energy ("self-sufficiency in 1980") have all exhibited this style so that policymaking in the United States has now adopted the Latin American style of rushing in with impulsive pseudo-solutions” (Hirshman, 1975). 30 Because we assumed from the start that the research-policy nexus is not linear, our questions to PM did not intend to relate national research outcomes and policy-making beyond their identification of influential studies. The answers did not allow a clear distinction between studies commissioned by government agencies and outcomes from research projects. Our sample of PM was too small to derive specific conclusions on different types of research use.

40 people’s consensus; some decisions entail society as a whole and need the confluence of society. Only social sciences can help us to progress in these fields.”

All PM considered that SSH research should cover more issues than the ones presently treated. The following topics and needs from SSH research were highlighted by several interviewees: More studies that assess and evaluate innovation policies and instruments: - “We need more rigorous evaluation studies… in terms of policy impact… We need good mechanisms to evaluate our instruments to progressively improve them. What matters most is to generate knowledge on which public policies work and which not… - “… South American countries are good to formulate and to implement but very bad for evaluating… There are no habits of transparency, of accountability and rendering of accounts, there is a problem of continuity in policy-making. There are various problems summing up, which on the whole are not stimulating for a systematic evaluation work… This is reinforced by the fact that the capacity to ‘accompanying’ institutions is limited when the institution that coordinates policies does not supply to the public the necessary information to proceed with the accompaniment.” “I believe an extensive social sciences research should be conducted on what exactly is needed in the country to design a good STI policy because PM don’t really know, neither producers, researchers nor people…”

A greater emphasis on recommendations in the innovation studies: - “… in this country, we constantly make diagnoses but on the question of how problems are solved I don’t see that researchers take a risk… The part that we need most as decision-takers is the identification of the possible pathways to solve a given problem, accompanied with the positive and negative features of each of them. - “… these studies often don’t go beyond the characterization of innovation dynamics.”

Social and hard sciences integration to tackle STI related problems: - “SSH research is one part of the research that must be taken into account in STI policy definition… For example, if I am going for energy policies, the whole scientific [hard sciences] research related to energy accounts maybe for 70%-80%; the rest is a global vision from the social sciences, more metascientific or meta-technical, i.e., beyond the technical part.” - “Among others, social sciences contributes with an external look on this historical debate between soft sciences and hard sciences, which has been associated to lobbies on ‘who gets funds for what’ and ‘giving funds to one kills the other’. A look from the social sciences incorporates the political dimension of science, the capacity to look at another dimension, beyond laboratories and mathematics; somehow it is the access point for society to seize the topic.” - “The classical example is the environment impact of industrial activities. ¿How do we face this problem? One needs to consider the whole problem, from a technical, scientific, health, etc. point of view, but also the social impact, the economic impact. The question thus becomes: How to integrate a working team that produces a study that allows one to tackle the problem? How do you create these ‘spaces of integration’ around relevant problems that demand knowledge from all these areas.” - “SSH research gives the north for what ‘science’ is going to solve. Without social sciences the horizon for the solutions provided by the rest of scientific research is undefined.”

Knowledge, social inclusion and development: - “Ten years ago we were concerned (and still are) with the linkage between research and production; now we are equally worried about the linkage between knowledge creation and social integration. In the same way as scientific research does not generate innovation on its own, innovation by itself does not necessarily improve life conditions. So, if we want innovation getting inserted in society, we need knowledge from social sciences, including maybe anthropology, on the one hand to get S&T development reaching people adequately and on the other hand, to research on the marginalization phenomenon, the question of fragmentation and the best ways to improve society’s conditions.

41 Everybody knows that this is not only a problem of resources assignment, but we have incomplete −and little scientific based− knowledge on these processes. - “We are not working enough on social inclusion issues, on integral and sustainable development with social inclusion. We should not involve only economics but the rest of social sciences too. Some topics cannot be approached if it is not through a mix of engineering, sociology, psychology and mathematics. We are not used to team work.

Looking at the implementation stage: - “People studying innovation always refer to the fact that, whatever is intended in this field, brutal social obstacles will be faced: vested interests, groups entrenched in several aspects and ways, legitimate and illegitimate processes through which different groups typically intend to defend their status quo… Without knowing and defeating these resistances, for example at the local level, it sounds idealistic to think that a social group is going to share a common vision, objectives, etc., and probably many programs and actions will remain formal efforts, ending up in supporting things that would happen anyway and that are not triggering a new dynamic. The relevant point is: ‘What is going to motivate the behavioural change that is needed to have society really engaged in innovation, in working together?’ The study of these dynamics, these resistances, these problems surging from different groups in a social context is very proper of social sciences … - “I think the discussion in LA of the innovation policy vis-a-vis the development policy is essential. There are advances, there is some discussion… *but+ there is a bias toward economics. Authors from developed countries are facing this discussion now. Lundvall himself now recognizes the importance of the social and the environmental questions. Except that there is a big difference between recognizing that these areas must be considered in innovation policy design and actually making this operational. How to coordinate and institutionalize this. In the case of LA, the social issue is very serious and very little treated by the innovation policy… The social, innovation and environmental questions must be treated together, as a whole. The discourse exists but the difficulty lies in the implementation stage. (…) There are institutional difficulties to make the articulation. I believe the SSH studies on innovation policy should evolve and look at the institutional environment.”

Thus, the task PM assign to SSH with regard to policy development is huge and seems to go well beyond the content of the studies they presently have access to. In spite of this, very little was mentioned in terms of joint efforts between PM and researchers to define a research agenda, even less so on the involvement of other actors –an issue that, in contrast, some researchers expressed as an essential need−. Of course, our limited sample of interviews might not be representative of PM’ view in general on agenda fixing. Actually, in the recent past different Latin American countries have carried out collective exercises to establish priority areas or sectors (hence, implicitly, research priorities) but there seems to be little evidence of this kind of experience focusing on social sciences (cutting across sectors) in support to development. 2.4.2

Knowledge inputs used in STI policy design: PM’ versus researchers’ view “There is nothing a government hates more than to be wellinformed; for it makes the process of arriving at decisions much more complicated and difficult.” J.K. Maynard Keynes (1936)

In 2009, a regional study was conducted to inquire on the use of innovation surveys in policymaking, based on interviews to decision-makers in Argentina, Chile, Colombia and Uruguay (Baptista et al, 2010).31 The interviews also co-laterally inquired on the other sources of information used by decision-takers in the design, re-design, and follow-up and evaluation of STI policies. Regarding the latter, the study presents the following conclusions (Ibid: 16-18): Decision-taking is based on a heterogeneous set of information sources, the importance of which varies depending on the stage of the PMP (Table 2.3). 31

The study was based on 36 interviews to decision-takers from 25 STI related institutions, such as STI and other ministries (planning, economy, industry); STI agencies and councils; development banks; etc.

42 A main source −in all four countries and at all stages (design, re-design and evaluation)− appears to be internal information of the institution, on the management of previously implemented or ongoing programs. A second, important source are reports from consultancies commissioned directly by the decision-taker or its institution, where information has been processed and contextualized according to the specific demand. These reports, together with strategic documents on STI policies, are extensively used in the design and re-design stages, and less so in the follow-up stage.

Table 2.3 – Sources of information used for decision-taking according to the stage of the PMP (design of policy and instruments; follow-up and evaluation; re-design)

Source: Reproduced from Baptista et al (2010: 17).

In the case of Uruguay, several interviewees referred to their personal experience as one of the main information sources for policy design and re-design. International consultancies and research are another privileged input. Especially in Chile, international information on global trends in different sectors and topics, as well as international consultancies, are particularly used for policy design. In contrast, statistical information is a less valued input by decision-takers, especially national innovation surveys results, though interviewees mentioned the use of data issued by international organizations, such as the World Bank, the World International Forum, Eurostat, the National Science Foundation, or the regional STI data network, RICYT. The report concludes that decision-taking is mainly based on the institution own internal and external (commissioned) information, which could reflect difficulties in accessing or trusting other national knowledge sources, as well as institutional articulation problems. In our study, we also inquired on the types of inputs used in the design of STI policies and instruments, according to both PM and researchers’ view. We presented a list of 13 possible sources

43 of information to interviewees, including an “other” option, and asked them to identify and rank the five inputs they consider are actually taken into account in the PMP in their country. The outcome must be looked at with cautious, in the first place because a few interviewees found it difficult to rank the items, which meant that they regarded, for example, three items of equal importance (in this case, the three items were assigned a 1 and the following, a 4). Also, as was already mentioned, our sample –especially with regard to PM− is too small to be representative of the STI political sphere in the selected countries. Finally, some interviewees did not answer this question, which explains that our total sample is smaller than the number of interviewees. Nevertheless, we treated the answers to this question as quantitative data because it helps to identify trends and a certain hierarchy in the utilization of policy inputs. Table 2.4 was built only with the answers from interviewees working in LA, i.e., in Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Mexico, Uruguay and Venezuela. The first part of the table shows the outcome when including answers from both researchers and PM; the second part draws on researchers answers exclusively; and the third part, only on PM answers. Looking globally, it is noteworthy that the most often selected item is “Personal knowledge and experience of the policy-makers” (column A), as was already highlighted for Uruguay in the above mentioned study. Almost 70% of interviewees marked this item among the five most important and 38% marked it as one of the two most important (column D). The breakdown between researchers and PM shows that this preference is somewhat stronger in the first group than in the second one, though in the latter case it still ranks high (third place) among the five most often selected items. Actually, more than a striking feature this is the acknowledgement of a fact of life. As Weiss once noted: “Policy makers have considerable knowledge of their own, based on their years of experience in the field, their education, the information and advice they get from everyone from their staff aides to their barbers and teenage children, the grapevine and the rumour mill, and models of policy from other jurisdictions and the effectiveness of policies those places have adopted. They are also inevitably subject to influence by the mass media, the books they read, and all the other sources of information that bear down on all of us. One research study is not necessarily going to sway their beliefs and make them see things a different way. Nor should it. In the last analysis, they have to use all their sources of knowledge and exercise their own judgment” (Weiss, 2003). Of course, this seems more in tone with PM’ view than with researchers’ opinion. The latter sometimes qualified their selection of this input with comments varying according to the country, for example: - “In this country we are all inspired!”, or “a decision taken by a PM is much more based on what he thinks than on studies”, and “personal knowledge and experience, yes, but presently it has to be understood as ‘lack of’ knowledge and experience”.

The second most cited input –considering the complete sample- is “Outcomes of deliberations between researchers and policy-makers”. This breaks down in a first rank for PM (with the highest mark as one of the two most important inputs) and a third place in researchers’ view. A researcher specified: - “Here I’m not referring to scientists specialised in S&T policy but rather to physicists, chemists, biologists, engineers, researchers in agrarian sciences, I mean, with the knowledge of their specific field.”

44

Table 2.4 – Inputs taken into consideration in STI policy-making in selected LA countries 1. According to interviewed researchers and policymakers (n = 39) INPUT OPTIONS

A*

B*

C*

D*

Personal knowledge and experience of the policy-makers

27

69%

15

38%

Outcomes of deliberations between researchers and policy-makers

23

59%

15

38%

Sector-based studies and diagnoses, and their policy lessons

19

49%

8

21%

Working lines or financing of international organisations

19

49%

13

33%

Analysis or conclusions from committees on specific issues

18

46%

7

18%

Budget negotiations at the national level (resources assigned to STI)

16

41%

7

18%

Personal or political interests of policy-makers

14

36%

12

31%

Pressures from advocacy groups, interest groups or lobbies

12

31%

6

15%

Outcomes of action projects carried out by the government or NGO

9

23%

3

8%

Publications of research outcomes (any science)

9

23%

4

10%

Quantitative data from surveys or similar

8

21%

5

13%

Others

6

15%

3

8%

Prospective studies

5

13%

1

3%

Public opinion

2

5%

0

0%

* A: number of times the option was marked as one of the five most important inputs. B: percentage of A in relation to the sample number (39). C: number of times the option was marked as one of the two most important inputs. D: percentage of C in relation to the sample number (39)

2. According only to interviewed researchers (n = 25) INPUT OPTIONS

A*

B*

C*

D*

Personal knowledge and experience of the policy-makers

18

72%

10

40%

Working lines or financing of international organisations

15

60%

11

44%

Outcomes of deliberations between researchers and policy-makers

12

48%

7

28%

Analysis or conclusions from committees on specific issues

11

44%

2

8%

Pressures from advocacy groups, interest groups or lobbies

10

40%

5

20%

Personal or political interests of policy-makers

9

36%

8

32%

Sector-based studies and diagnoses, and their policy lessons

7

28%

2

8%

Budget negotiations at the national level (resources assigned to STI)

7

28%

2

8%

Quantitative data from surveys or similar

4

16%

3

12%

Publications of research outcomes (any science)

4

16%

2

8%

Others

3

12%

2

8%

Outcomes of action projects carried out by the government or NGO

3

12%

1

4%

Public opinion

2

8%

0

0%

Prospective studies

1

4%

0

0%

* A: number of times the option was marked as one of the five most important inputs. B: percentage of A in relation to the sample number (25). C: number of times the option was marked as one of the two most important inputs. D: percentage of C in relation to the sample number (25)

45 3. According only to interviewed policymakers (n = 14) INPUT OPTIONS

A*

B*

C*

D*

Outcomes of deliberations between researchers and policy-makers

11

79%

8

57%

Sector-based studies and diagnoses, and their policy lessons

11

79%

7

50%

Personal knowledge and experience of the policy-makers

9

64%

5

36%

Budget negotiations at the national level (resources assigned to STI)

8

57%

5

36%

Analysis or conclusions from committees on specific issues

7

50%

5

36%

Outcomes of action projects carried out by the government or NGO

6

43%

2

14%

Personal or political interests of policy-makers

5

36%

4

29%

Publications of research outcomes (any science)

5

36%

2

14%

Quantitative data from surveys or similar

4

29%

2

14%

Prospective studies

4

29%

2

14%

Working lines or financing of international organisations

4

29%

2

14%

Others

2

14%

1

7%

Pressures from advocacy groups, interest groups or lobbies

2

14%

1

7%

Public opinion

0

0%

0

0%

* A: number of times the option was marked as one of the five most important inputs. B: percentage of A in relation to the sample number (14). C: number of times the option was marked as one of the two most important inputs. D: percentage of C in relation to the sample number (14) Source: Based on 39 interviews to researchers and PM in the field of innovation in LA, 2009-2010.

Thus, it appears that two main inputs in policy-making consist in knowledge embodied in people, which obviously includes tacit forms of knowledge. However, PM also assign the highest rank to the following type of explicit, ‘written’ knowledge: “Sector-based studies and diagnoses, and their policy lessons”. This contrasts with their much lower ranking of “Publications of research outcomes (in any science)” in spite of the fact that, presumably, sector-based studies are also often published as research outcomes. One way of interpreting this is that PM want processed, comprehensive and clear-cut sector-based information, that makes their decisions easier and not more complex. The same argument could apply to the rather moderate ranking of "Quantitative data from surveys or similar" according to the observation of a PM, cited in the mentioned study on innovation surveys in LA (Baptista et al, 2010): - “… innovation surveys have a certain value, (…), some data are collected, (…), they give some numbers about the movie but they don’t show you the movie, and the movie runs through very complex things, through details, through a network where macroeconomic aspects combine with personal circumstances, technological realities. There is a need to go into detail and, in this sense, sector-based studies shed other points of view, the same [occurs] with reflections of specific research projects on these topics.”

Sector-based diagnosing is indeed an exercise that is repeated time after time (or government after government) in many LA countries. Any PM wants a fresh diagnosis of the economy’s sectors or branches as a frame for his actions. Possibly, PM would equally value “Prospective studies” but, since LA has a severe deficit in this field, it does not (and could not) appear as an often cited input. On the researchers’ side, the second most cited item is “Working lines or financing of international organisations” (60%). In most LA countries, financing from IDB, World Bank, EU, etc. is essential in the field of STI and some researchers associated this influence with the fact that STI policy instruments are very similar in many countries. In contrast, only 29% of PM considered this item as one of the five most important in policy-making. A few PM categorised a specific type of influence of

46 international organisations in the “others” item, namely publications in the field of STI that are used not only to get acquainted with new concepts and tendencies but also to get information helping benchmarking exercises or comparative studies. “Analysis or conclusions from committees on specific issues” appears as a relevant input in policymaking, both in PM and researchers’ view. This reinforces the importance of specific, processed knowledge, transmitted through (trusted) people. “Personal or political interests of policy-makers” was confirmed on both sides as an influential factor: it is not one of the most often cited inputs (36% of the whole sample) but, when it is, it is almost exclusively ranked in the first or second place. “Pressures from advocacy groups, interest groups or lobbies” received a different treatment in each group of interviewees. PM mostly disregarded it as an input, while 40% of researchers marked it as one of the five most important inputs. According to researchers’ comments, this difference in perception can be mainly attributed to the already mentioned fact that, in several of the surveyed countries, PM heading the main STI agencies or ministries stem from the [hard] science community. Thus, our interviewees, who work in social sciences, view the scientific community as a strong lobby in STI policy-making. The following comments are explanatory: - “Interest groups pressures? In this case, it comes basically from the so-called scientific community”. - “The most important item is ‘Knowledge and experience of PM themselves’, which is the scientific community. Because, here, all policymakers are part of this community. So, ‘personal interests’? ‘pressures of interest groups’?, in the case of STI policy the interest group is the hegemonic actor itself, do you realise? I mean, in other public policies there are interest groups and lobbies who put pressure on PM: physicians influence the health policy but they are not the ones who make it; army officers influence as a lobby the national defence policy but they are not the defence policy. In the case of STI, the lobby itself is who makes the policy; it is something even more powerful than lobbying.”

“Outcomes of action projects carried out by the government or NGO” 32 was included as an option in the questionnaire so as to observe whether PM use the lessons from non-research based projects, for example to design or improve policy instruments. The influence of this item in policy-making appears to be moderately high in PM's view (43% of PM mentioned it as one of the five main inputs but rarely as one of the two first) and low in researchers’ view. “Public opinion” is practically inexistent as input. At the most, some interviewees referred to a mounting influence of this factor: - “Public opinion has an increasing influence; it is very thin though important. Public opinion surveys are conducted and published, etc. I believe society shows some interest.”

As was said, “Quantitative data from surveys or similar” is not highly ranked as a policy input by PM and even less by researchers. In the previously mentioned study on innovation surveys in LA (Baptista et al, 2010: 33), the main obstacles PM reported for using these surveys are inexperience and lack of analytical capacities. The deficient articulation between actors of the NSI and the limited participation of PM in the design of surveys are the most often mentioned obstacles for supply to match demand. The study confirmed that, the greater the participation of PM in design, the more they use it. Strangely, there is no mention of the need to involve innovators themselves, and more generally, firms’ managers and firms’ associations, in the design of surveys questionnaires. This is a question we will come back to in chapter 3.

32

For instance, in Uruguay two large projects have been carried out by government agencies since the middle of this decade to promote competitivity through support to clusters' and networks' building in a variety of sectors. They are a source of knowledge, among others, on local capacities at the firm level.

47 The countries analysed in this study stand at different stages of STI policy development. Taking extreme cases, Brazil has a long tradition and many specialists with an extended trajectory, both on the research and the policy-making side, while in Costa Rica innovation policy and research are in their infancy. Thus, one could think a priori that the type of inputs might differ among countries. Indeed, there are some specificities at the country level but they do not alter significantly the previous analysis.33

2.5

Strong and weak nexus: an illustration

2.5.1 Local productive and innovative systems in Brazil: an example of good match between researchers and policy makers The Portuguese term "Arranjos Produtivos Locais" or APLs, i.e., local productive and innovative systems (LPIS), was proposed during the development of a regional research project, called “Globalisation and Localised Innovation: The experience of Local Systems in MERCOSUR Countries and Policy Proposals for Science and Technology”. This two-year project started in 1997 and was led by Helena Lastres and Jose Cassiolato, both researchers at the Institute of Economics, Federal University of Rio de Janeiro, Brazil. This project was the opportunity to launch an extraordinarily successful research network, RedeSist (http://www.redesist.ie.ufrj.br/Ev/home.php) actually including near thirty universities and research centres all over Brazil. Soon after the conclusion of the project, the concept of LPIS was “academically” launched (Cassiolato et al, 2000). RedeSist define LPIS as follows: “Local productive and innovative systems are territorial agglomerations of economic, political and social agents, focusing on a specific set of economic activities and presenting relationships between them, eventually incipient. They generally involve the participation and interaction of firms, be them producers of final goods and services, suppliers of intermediate goods and equipment, consultancy and services firms, firms devoted to commercialization, clients, etc., and their diverse forms of representation and association. LPISs include as well a diversity of other public and private organizations committed to training of human resources, like technical schools and universities; research, development and engineering; policy making, promotion and financing (Ibid).” The importance that the concept, born in academic circles, acquired in policy-making can be traced back through several indicators. One particularly telling is the formation in 2004 of a Permanent Working Group (PWG) on LPIS, aiming at coordinating governmental actions to give integral support to LPIS. In June 2010, the renewal of people belonging to the PWG was announced in the Official Newspaper, under the heading of the Ministry of Development, Industry and Foreign Commerce (http://www.mdic.gov.br/arquivos/dwnl_1290022992.pdf).Thirty three institutions belong to the PWG, including some of the most important and powerful ones in Brazil. Table 2.5 organizes them by type of institution.

33

For example, “Personal or political interests” was not often mentioned in Chile, Colombia and Uruguay; and “Working lines or financing of international organisations” was not mentioned in Brasil. In contrast, “Personal knowledge and experience of the policy-makers” was pointed at in all countries.

48

Table 2.5 – Institutions belonging to the Permanent Working Group on LPIS

Ministries

Banks

National Agencies

Public Research Institutions

Private organizations

                                

Development, Industry and Foreign Commerce Planning, Budget and Management Work and Employment Tourism Mining and Energy Education Agrarian Development National Integration Agriculture and Husbandry Environment Science and Technology Economy National Bank for Economic and Social Development Bank of Brazil Bank of the Northeast Bank of the Amazonian Bradesco Federal Saving Bank National Institute of Metrology SEBRAE (Brazilian Support Service for SME) FINEP (Financing Agency for Studies and Projects) CNPq (National Council for Science and Technology) National Council of Regional Secretaries for STI SENAI (National Service for Industrial Training) National Agency for the Promotion of Exports EMBRAPA (agricultural research institution) Institute for Technology Research IPEA ( Institute for Applied Economic Research) Free Zone Manaos Development Company of San Fernando Valley and Paraiba National Confederation of Industry Institute Eduardo Lodi Movement Brazil Competitive

Source: Based on http://www.mdic.gov.br/arquivos/dwnl_1290022992.pdf

The last survey of LPIS, performed in 2005, identified 957 of them. The following map gives account of them in the whole country. According to the same source, the following table can be made, combining several productive activities of the LPIS and their geographical locations (Table 2.6). One of the high-tech LPIS, in biotechnology, is located in the most industrialized region of Brazil, the Southeast region.

49 Table 2.6 – Productive specialisation of the LPIS by regional location  NORTH (States of: Amazonas, Roraima, Amapá, Pará) Wood and furniture; furniture

Ornamental stones

 NORTH-EAST (States of: Maranhao, Piaui, Ceara, Bahia, Natal, Paraiba, Pernambuco, Alagoas, Sergipe)  CENTRE-WEST (Mato Grosso, Goias, Mato Grosso do sul)  SOUTH-EAST (Minas Gerais, Espiritu Santo, Rio de Janeiro, San Paulo)  NORTH-EAST

Diary

 CENTRE-WEST, NORTH-EAST, NORTH

Leather and Shoe-Making

 CENTRE-WEST, NORTH-EAST, SOUTH-EAST, SOUTH (Santa Catarina, Rio Grande do Sul, Paraná)

Garment

 CENTRE-WEST, NORTH-EAST, SOUTH-EAST, SOUTH

Metal -Mechanic

 NORTH-EAST, SOUTH-EAST, SOUTH

Fish Farming

 CENTRE-WEST, NORTH-EAST, NORTH

Precious stones and jewels

 SOUTH-EAST, SOUTH

ICT

 NORTH-EAST, SOUTH-EAST

Apiculture

 CENTRE-WEST, NORTH-EAST, NORTH

Popular Crafts

 NORTH-EAST, NORTH

Fruit farming

 NORTH-EAST, NORTH

Source: Based on http://www.mdic.gov.br/arquivos/dwnl_1290022992.pdf

Particularly interesting is how the term LPIS has permeated the whole Brazilian economy, even far away from official denominations. For instance, the LPIS of electronics, in a specific location of the state of Minas Gerais, Santa Rosa de Sapucaí, presents itself as such when looking for partnerships or when performing advertising. Most probably, those that self-denominate as the electronic LPIS of Minas Gerais do not know where the term comes from, showing how much the denomination has been integrated into the “common sense” of regional industrial development.34 The term has gone further, to reach the regional policy-making. Between 2005 and 2007, for instance, SEBRAE in association with ECLA and the Inter-American Development Bank, organized several training courses for LPIS managers (Teixeira and Ferraro, 2009). Given the seminal work of Piore and Sabel and of Porter on clusters and economics of agglomeration, a puzzling question is why the term LPIS came to be recognized –and not only in Brazil35– as the way to refer to the organization that helps gaining competitiveness from a widely understood experience of cooperation. Teixeira and Ferraro suggest that a main reason is that their proponents, especially Cassiolato and Lastres, put a strong emphasis on the specific historical conditions of peripheral countries, making the concept more earth-to-earth for concrete policy work (Ibid: 17). 2.5.2

SSH research and STI policy-making in Venezuela36

SSH research in STI Evidence from different sources tends to reveal a certain weakening of STI research in Venezuela as 34

More information on the electronics LPIS in http://www.valedaeletronica.com/index_arquivos/elvalle.htm In 2004, SEBRAE, ECLAC and the German GTZ organized in Brasilia a Latin American Workshop on LPIS. The final report of this workshop can be downloaded from: http://www.eclac.org/ddpe/noticias/noticias/2/15512/InformeTallerSPLBrasilia2004.pdf. 36 The following case study is a contribution from Ignacio Avalos, sociologist, profesor at the Universidad Central de Venezuela. 35

50 compared to a couple of decades ago. On the one hand, several research groups have lost strength and few new ones have come to existence. There are fewer researchers in this field, and the upcoming of a new generation is not clear. On the other hand, there seems to be a greater relative strength (or a less pronounced weakness) in the field of science history and sociology than in innovation topics related to public policy. Innovation studies are often micro-based, firm-oriented and sometimes commissioned by firms, and firms, and mainly –though not exclusively– related to technology management. Usually, these analyses do not have implications for policy development. There appears to be a divorce between the issues of interest to researchers (to some extent researcher’s own history might hinder his opening up toward other topics) and those that decision-takers in the public arena would eventually need. The necessary communication between researchers and PM in order to make out a joint working agenda of reciprocal convenience is missing. The present political atmosphere is probably largely accountable for this, as will be explained below. It is fair to notice, though, that STI studies have always tended to be on the margin in Venezuela (with the exception of some notorious cases of direct or indirect connection and impact) but the situation has worsened in recent times. A significant event has been the passing of the Organic Law of STI (LOCTI) in 2006, which established that firms must allocate between 0,5 percent and 2 percent of their gross income (depending on the nature of the firm) to STI activities. The firm has the choice to spend these funds either in internal activities or in contributing to third parties. In the latter case, some research centres have obtained additional resources, above their regular budget, which have been used to carry out studies coinciding with their working line and, in a less proportion, to attend demands of the industrial sector according to the latter interests. Within this context, two present trends can be visualized. On the one side, the oldest research groups on STI, almost all located in universities, work with a different political view than the government one (in spite of being public universities) and are, in some way, considered to be on the ‘opposition’ side. Therefore, their services are little demanded. They also suffer from the general budget restrictions that have affected universities. However, as was said, for some of them LOCTI was an important source of revenues. On the other side, STI studies carried out in the frame of government related institutions (academic centres; Ministry of Science, Technology and Intermediate Industries’ agencies; National Science and Technology Fund, FONACIT) have a strong ideological base. They usually claim that science and technology must be at the service of the setting up of the XXI century socialism and that the traditional capitalistic view of STI is incompatible with the new proposal for society. There is no diffusion strategy of research outcomes, beyond the classical publication of findings in books and journals and their discussion in academic seminars. Only in few occasions an effort can be noted to ‘knock on the door’ of the government with the purpose of influencing public decisions. Finally, there is a notoriously scarce linkage between the different research groups, and weak nexus with institutions of other countries or international organizations. Summing up, SSH studies on STI are in crisis in Venezuela. Naturally, a more rigorous view of the situation would require a detailed inventory of capacities. The STI policy-making process On the public policy side, the deficit of studies and data does not seem to be viewed as a serious problem. Policy design is usually fed with the following three elements: The construction of the XXI century socialism as an ideological desideratum. Concepts such as endogenous development, technological sovereignty, knowledge democratization, knowledge dialogue, diversification of technology transfer sources for geopolitical reasons, etc. mark the conceptual pathway for STI policies.

51 The literature that emerged during the 1960’s and 1970’s from the Dependency Theory, which inspires and explains a good deal of the present STI policy. A certain ‘common sense’ that came into shape through time, mainly in the 1980’s and 1990’s, derived from the literature on the social nature of the innovative processes (which capacities matter, which actors participate and how they interrelate, how networks build up, etc.). LOCTI is a clear example of this influence as well as some other programs of the Science and Technology Ministry, such as the ‘Socialist Innovation Networks’. In these and other cases, the government pretends to assume the organizational and functional aspects of innovation according to the concept of National Systems of Innovation, though not clearly successfully as will be further argued. Summing up, it could be asserted that −at least at the declarative level and omitting some exceptions− the Venezuelan STI policies are guided by the objective of making way to a socialist society, are grounded in a somewhat renewed Dependence Theory, and are fed at the operational level by evidence provided by the international literature on innovation processes. Policies are not developed through consensus but rather reflect the government’s point of view. The political conflict, which has divided the country in two, affects the making of the social tissue that is needed for innovation systems to exist and work well. A substantial part of the business sector is suspicious about government and something similar occurs in the universities, which host a good deal of the national research capacity. In this context, trust and cooperation relationships hardly arise in spite of being essential for the innovation processes. Indeed, as evidenced by the specialized literature, building trust is a complex phenomenon, highly influenced by the quality and intensity of the relations between actors with different capacities and interests. The approval of a policy or an instrument rests essentially on the extent to which these are ‘socialistoriented’; their relevance or feasibility in terms of the development of innovative capabilities is of lesser concern. Therefore, there is no real need to back them up by research. There are many examples in this regard. The most important one is, undoubtedly, the mentioned LOCTI. This new and very relevant instrument is a key piece of the institutional architecture that pretends to govern STI activities in Venezuela. It now appears that the articles of the law referring to the contribution of the business sector will be modified. The motivation is not clear since this issue is discussed exclusively at the highest bureaucratic level. However, a hypothesis has been spreading in the sense that LOCTI’s norms regulating the contribution of the business sector are not applied according to a ‘socialist code’ but rather to a ‘production code’, i.e., in the firms’ interest. And this –goes the argument− is a misunderstanding of the official policies, oriented toward the construction of the XXI century socialism. 37 There is no evidence of any rigorous evaluation study of LOCTI implementation (neither inside the government nor commissioned to some institution or research centre) that would determine its real impact and eventually back up the law transformation. According to the Ministry, LOCTI has meant an STI investment increase that situates this spending around three percent of GNP. More generally, many high-level government officials have qualified the lack of studies evaluating policies and instruments as a serious deficit. A common complaint among researchers and high level civil servants concerns the excessive centralization of decisions concerning public policies, in the field of STI and more broadly. ‘Misión Ciencia’ (Science Mission) is a case at stake. This ambitious program, which was short-lived and had a very limited impact, was launched without a basic substantiation (concepts, data, etc.) except for the presidential intention of democratising knowledge generation and use.

37

The anual funds stemming from the business sector in fulfillment of LOCTI are roughly ten times more than the total resources allocated in the national budget to FONACIT.

52 At the same time, there are no institutional mechanisms for the systematic search of national and international studies that could contribute to the understanding of STI related issues and provide some support to public decision taking. Finally, during the 11 years of the present public administration, the rotation of high-level authorities in the field of STI has been particularly high: at least seven ministers have been succeeding and a similar number of FONACIT presidents. Naturally, these changes at the highest level often entail the substitution of civil servants at the middle level and the consequent discontinuity in management. This exacerbates the problem derived from the lack of an educated bureaucracy in the field of STI, which in turn limits interaction with the research centres. Many researchers view these limitations as a serious obstacle to the establishment or strengthening of nexus with the public sector.

53

3.

Conclusions and policy recommendations

In this chapter, we first present some broad conclusions from the study carried out. Then we look thoroughly at what we perceive is a crucial issue for fostering the research-policy nexus, namely, the demand-side problem of the relations between SSH research and STI policy. 3.1

Overall conclusions from the empirical work On researchers’ side

1. All researchers assert that contributing to the PMP is one of their group’s objectives, and they all reported some influence of their work on policy development. For sure, at all times, research has been presumed to produce knowledge, and knowledge should affect what governments do. But some decades ago, when big surveys started to be carried out on ‘research utilization’, there was little recognition that PM had also knowledge from other sources, and researchers tended to presume that findings should have an immediate influence on policy. Weiss, who was already working on research impacts at that time, says that many researchers, especially evaluators, believed their purpose was basically to show decision makers which policies and practices to follow (Weiss, 2003). Our interviews show that the departure from this linear view toward more refined and complex nexus models is not exclusive of the scholars devoted to researchpolicy issues. Though reflecting on discernable impacts of their research does not seem to be a usual practice of our interviewees, they were all well aware that much else besides research affects policy decisions, and that the influence or impact they might have is, to say the least, uncertain. 2. Overall, the main types of influence they perceive are conceptual (influencing the way of thinking of PM) and ‘embodied’ (through the movement of persons between the academy and the policy spheres). 3. Some instrumental uses and impacts of their research on the PMP were identified. However, once considered all together, these reported impacts seem to understate the importance, quality and diversity of the work carried out by the innovation research groups. This confirms of course the nonlinearity of the research-policy nexus but it also reveals the extent of the space for a better use of all SSH has to offer from an innovation perspective. 4. Instrumental uses of research outcomes have mainly resulted from: Collective exercises at the regional level (Bogota Manual; RICYT’ STI indicators), but practically no other examples of influential regional initiatives or networks were reported. Specific demands from PM (studies commissioned by government agencies), which tend to be punctual rather than agenda-based (consultancy work prevails over more in-depth research). However, in several countries research groups (or persons) have been called to participate in preparatory workshops aimed at defining the national STI policy or strategy. It seems that in these cases there is a mix of conceptual and instrumental influences. 5. In brief, researchers’ perception is that the impact of their research is not clear-cut; it is mainly intangible, built up through time and through many actors; and it is highly dependent on the particular institutional and political context of the moment. Strictly in this regard, our empirical work could look as 'a painful reconstruction of the obvious', since at the end of the 1970s Weiss already stated: “The process is not one of linear order from research to decision but a disorderly set of interconnections and back-and-forthness that defies neat diagrams” (Weiss, 1979). However, several interviewees noted that they did not know of any study in the field of STI that focused on this process in LA from a social sciences perspective.

54 6. Besides what could be called straight convergences among researchers’ perceptions, the empirical work also showed some ‘revealed’ convergences, this is, convergences that do not consist in researchers saying similar things but in their expression of diffuse opinions that the analysis shows to be referring to similar issues. An example of this is how researchers view PM. In different ways, they say that the behaviour of PM are strongly influenced by the institutional setting, mainly by the layers of institutions involved in one way or another in innovation initiatives. The extent to which researchers’ information, analysis and recommendations are taken into account at policy level is related to this. A ‘revealed’ convergence would then be that the nearer the researchers are to the PM (nearness can be geographic or sectorial), the more the latter will rely on the work done by the former. 7. Among the divergences that appeared from the interviews, it is worth noting researchers’ view of the motivations and effectiveness of PM. For several innovation researchers, PM have an a priori agenda, relatively immune to research results, in some cases to such an extent that they go on with policies that research outcomes have explicitly shown to be misplaced. Other researchers have a more positive opinion of STI PM, indicating that PM implement policies that enhance innovation results and empower innovation actors. An important question is where these divergences come from. Besides reasons derived from differences at the country level, positive opinions are usually related with reporting good personal communications between researchers and PM as well as ‘cognitive nearness’ between them, for instance for belonging to the same economic school of thought. Negative opinions correspond approximately to institutional and cognitive distance. 8. Divergences can also be found on the issues researchers think bear more influence on the PMP, which is to be expected because influences are strongly contextual. The same happens with the main problems researchers identify for more effective policy-research links. There is a wide-ranging diversity in researchers’ view on the obstacles to the research-policy nexus. We classified them in five categories: (i) mismatches between research supply and demand, or related to the so-called ‘two communities problem’; (ii) limitations of research itself; (iii) obstacles derived from the mere nature of the policy-making process; (iv) governability issues; and (v) external factors. 9. As concerning theoretical contributions of the innovation research groups, on the one hand interviewees referred to a diversity of innovation related concepts (sometimes of their authorship) that they contributed to diffuse and/or adapt to the Latin American context. On the other hand, the following theoretical approaches are worth highlighting: - The already mentioned Bogota Manual is an example of a theoretical approach intended to influence STI policy. It was the product of a collective effort to develop indicators able to better assess what firms do in developing countries in terms of innovation and how well equipped firms are to foster innovation as part of their business strategy. Theoretical approaches of the sort are important because they illuminate both the search and the analysis carried out. Einstein phrase (actually referring to a discussion on quantum physics) is fully valid here: “It is the theory that decides what we can observe”. Innovation theories with a more global social turn than mainstream economics have been developed in highly industrialized countries for decades; the modifications they need to give account of Latin American realities require theoretical approaches such as the one referred to. - Another interesting example giving account of efforts to picture and analyse the innovation situation on which STI policy needs to operate, is Brazil’s work around Local Productive and Innovative Systems (LPIS). It entailed a huge empirical work, covering the whole country and tenths of sector-based studies in natural resources, agriculture, manufacture and extremely diverse services. The point is that the empirical work was carried out from an original theoretical perspective: though related to the cluster concept, it departed from it in several aspects so as to capture forms of innovative relatedness between actors and activities that the more formal cluster theory cannot make visible in realities like the Latin American one.

55 10. However important, these collective efforts at the regional and national level are exceptions rather than the rule. Overall, there is little tradition of joint or complementary work between innovation groups in SSH. Researchers know each other's and read their mutual publications, but not much more than that. For a variety of reasons (including sometimes rivalry, for example derived from unequal access to information, and personal reasons) the research groups are relatively disconnected, each one working on some facets of the innovation processes. Connectivity occurs preferably at the international level: researchers look for their peers in their specialisation field and establish their networks through international forums or joint research projects. In short, knowledge tends to accumulate in different fields of STI with apparently few mechanisms to integrate and articulate these different ‘pieces’ of knowledge towards the building of a Latin American vision of development that would distinguish between the distinct realities of the countries in the region. On PM’ side 11. The task PM assign to SSH research is impressive when compared to the use they report of published research results. To a certain extent, this indicates that much more effort is needed to ensure that project findings inform policy-making in a useful way. Among others, PM demand studies that go beyond the characterisation of innovation dynamics to include more recommendations on concrete ways to tackle the identified problems. Also, more studies that assess and evaluate innovation policies and instruments. Obstacles of this kind to the research-policymaking nexus were clearly identified in a recent survey to European PM, senior advisors and knowledge transfer specialists (European Commission, 2008: 13): "Policy-makers are focused on the need to bring practical solutions to particular policy-development issues. They need information that will inform their decision-making process, either ex ante in defining policy or ex post in evaluating policy choices. This information must be accessible, politically useful, and contribute to finding practical solutions to problems. (…) [Researchers] also need to understand the importance of translating their research findings into policy useful material, and the importance of supporting policy-makers in identifying appropriate solutions to problems". 12. The integration of hard and social sciences is seen as particularly important to analyse STI related development problems in their different dimensions, among others marginalization and social exclusion issues. 13. In spite of their belief that SSH research should tackle many important issues, there was practically no mention of the need or convenience to work collectively on agenda setting. Beyond the shortcomings in the ways research results are communicated, our study clearly shows that inputs used in policy-making are related to knowledge embodied in people: PM rely heavily on “outcomes of deliberations between researchers and PM” as well as on their “personal knowledge and experiences”. On modes of articulation between SSR and policy in the field of STI 14. Based on the empirical work, we propose three modes of articulation between SSR on STI and STI policy-making: “arm’s length”, “hands-on”, and “connected distance”. This proposition derives from evidence provided by the interviews, without further independent institutional checking. The taxonomy is thus based on subjective appreciations from both researchers and PM on how their relationship established and how it evolved over time. The arm’s length mode, as its name suggests, implies that researchers work in a kind of ‘mode 1’ of knowledge production, following the well-known Gibbons et al (1994) concept. It is characterized by the non-negotiated setting of the research agenda. On the PM’ side, this implies that the PMP does not include a systematic consultation of research outcomes or of researchers themselves. The two spheres of action, research and policy, might have encounters, even planned encounters, but the logic that moves both spheres results in few points of contact. Maybe PM do take into consideration

56 research results but the national research community and PM work at a distance (thereby also avoiding conflicts). Researchers might eventually be hired by an innovation related agency to perform commissioned research, but the working agenda of the innovation research community will not be much influenced by PM needs, and research will be understated among the knowledge inputs used in the PMP. The hands-on mode implies that, by different routes, innovation research agenda and innovation policy design are strongly connected. These routes can be people moving from academia to policy and the other way round, carrying moving questions, demands, concepts and proposals. Other routes are made up of well-oiled channels of communication, through specific think-thanks, joint attendance to conferences, joint calls for research projects aimed at gathering needed information and analysis, effective communication of research results and recommendations, etc. Finally, the connected distance mode implies that, even when each community working agenda relies on its own logic, there are nevertheless bridges that connect them, albeit not in a systematic way. These proposed modes are only stylized idealization of current situations. Not only is it hard to find a concrete national case conforming completely to any of such modes but, also, concrete situations are not concordant with one mode all the time, eventually shifting from one mode to the other at different moments. As an approximation, Brazil, Chile, and also Cuba, could be viewed as examples of the hands-on mode; Venezuela could presently be an example of the arm’s-length mode, and Argentina, Uruguay, Colombia, Costa Rica and México, at different rates, can be considered examples of the connected distance mode. On innovative contexts 15. Context is an important part of the story, according to the international literature on research influences and impacts. Individuals in policy-making positions do not act alone or in isolation from everything going on around them. They are bound by the availability of resources, the constellation of interests affected, the line-up of supporters and opponents, and previous decisions. To what extent can convergences, divergences and modes of articulation be traced back to national innovative contexts? It seems reasonable to relate the ‘hands-on’ mode to a national agreement on the importance of innovation for the future of the country. This could be the Brazilian case. Even if it is not so clear for Chile, it is nevertheless true that STI is relatively strongly institutionalized in this country and, moreover, it has an important ‘independent’ source of STI policy financing coming from copper production taxes. Cuba is a different case, given that the hands-on mode is more the outcome of a purposeful building than the result of an evolving articulation pattern. Where strong conflicts characterize the relationships between the innovation research community and PM, sometimes as part of a more general conflict between academia and government, arm’s length relationships would be the expected mode. This seems to be the case of Venezuela. The connected distance mode can have different underlying motivations. In a country like Argentina, the relative weakness of the articulation between innovation research and policy-making can be partly explained by the historical high weight of natural scientists in the orientation of STI policies. In contrast, in Uruguay, such weakness should be explained by the relatively low priority given historically to STI: the presently growing political importance of the issue thus faces a weak tradition of dialoguing. One of the challenges ahead is to devise institutional tools to facilitate the move towards more hands-on modes of articulating SSH research and policy-making in the field of STI. Doing so implies changing the innovative context. On the measurement of influences and impacts 16.On this issue, we bring out the following observations:

57 - An important factor that adds to the difficulty of assessing research impact on policy-making is temporality. Research might influence PM’ ways of thinking but PM are only temporally decisiontakers. How much of this tacit, conceptual capital is institutionalised and survives in the following government or the next stage of the political process is an issue that would require in-depth country case studies, so as to get a better understanding of what should or could be done in this regard (beyond the obvious need for State rather than ‘government’ policies, and the unchangeable complex nature of the PMP). From our interviews, it seems, for instance, that Brazil would be a case where this kind of ‘institutionalisation’ occurred, while Venezuela presently appears as a case where the research-policy nexus is loosing ground. For sure, the long-term influence of research might be quite distinct from the short or medium term impact. As one researcher expressed: “To build up something serious, many years are needed and the question is how to position this issue in the public space, beyond one government and another; how to get the next government building on top of the good things from the previous one without saying ‘this is all useless’.” - In spite of the abundant international literature on the SSH research to policy link, there seems to be very few studies proposing methodologies on how to define and somehow measure which types of knowledge exchange between the research and the policy spheres are effective. As was shown in chapter 1, theory has rather focused on building models that reflect the complex journey of research to policy as well as the systemic interplay of various factors and actors. Both in developing and developed countries, case studies have been the way to generate or apply these models (in other fields/sectors than STI from a SSH perspective), once it was acknowledged, from previous empirical work, that fixed survey forms with closed-ended or multiple choice questions are not enough to capture the nature of the nexus. By ‘defining’ and ‘measuring’ effective links we mean, for example, exploring questions such as collective agenda setting or policy instruments’ evaluation. Why, how and with which benefits do some countries succeed, and others not, in involving (local or regional) research capacities in these issues? Alternatively, what benefits in terms of policy development and implementation arise in countries where a broad range of knowledge purveyors does play an intermediate role? If we had some types of indicators reflecting different stages, intensities, and benefits from knowledge exchanges between research and policy, the deficits of different countries in this regard would be more neatly pinpoints, stimulating a better attention to neuralgic points in the research-policy nexus. On the ground of these observations (and others that follow), it seems that there is a broad field for more STI policy research in LA. Several interviewees referred to the lack of integration of different branches of social sciences to progress in the study and understanding of the STI policy-making process. An economic-based (industry oriented) framework prevails, with limited interactions with political and public policy sciences, and other social sciences (sociology, history, psychology, law). In this regard, it is worthwhile to recall that, in the North, STI policy research is considered to fulfil its collective responsibility of helping construct more effective policies when the four components of this field are addressed: STI policy science, STI policy engineering, STI policy entrepreneurship, and STI policy scholarship (chapter 1, p. 18). 3.2

The importance of research policy demand for evidence-informed STI policies

Looking at the challenge of fostering a better matching between SSH research and policy, in the field of STI, let us first say that we are not implying that every outcome of SSH research should inform policy, or that STI policy should be mainly based on research. Both spheres must retain important degrees of autonomy to be effective in their own terms, and to the extent that such autonomy exists, tensions and some kind of mismatches are unavoidable. The real problem appears when,

58 rather than relative autonomy, disconnection occurs. 38 We therefore propose the concept of ‘connected autonomy’. Connected autonomy between SSH research on STI and STI policy requires the will of researchers to be connected to policy, as well as some expression of research needs on the PM side. Though no PM would probably express a lack of interest in SSH research, their perception that some work must be done to communicate policy needs to researchers is usually missing or, at least, as our interviews show, little action is actually carried out in this regard. In fact, the demand-side problem of the relations between SSH research and STI policy is similar to the demand-side problem faced by STI policy itself. Regarding the latter, Georghiou (2007:1-2) put it this way: “Imagine trying to cut a piece of paper with just one blade of a pair of scissors. It’s near impossible. Yet that is what we try to do with innovation policy. (...) Innovations are the product of the creative interaction of supply and demand. However, in focusing on how to increase the supply of innovative businesses, policymakers have lost sight of the importance of demand”. Now, in the context of our study on research-policy interfaces, cutting a piece of paper with only one of the scissor’s blades means defining STI policy without the insights from SSH research. PM have diverse means for steering STI research. It can be done through commissioned research, for instance, on the general social backgrounds of firms’ owners in a given sector where innovation is to be fostered. It can be done through indirect ways, by announcing specific policy goals that can be taken up as SSH research targets. Obviously, there are no warrants that research outcomes will be taken into account in policy design, but this is more probable to happen when research responds to explicit policy concerns. Thus, we are faced with a double ‘one blade scissors’. On the one hand, interviews confirmed that PM are in need of more information on demands from different actors of society that could be addressed through innovation policies and instruments, including, though not exclusively, demands from the production sectors. The translation of existing needs into explicit innovation demands is not a simple issue. On the other hand, this lack of knowledge does not only obstruct the design of effective promotional instruments, it also makes difficult for PM to define and communicate research needs to the academy. Research knowledge on both of the second scissor blades could inform policy; actually, researchers do work on these issues but too often in disconnection from the policy-making spheres.

38

This is an old problem in the realm of STI studies. University−industry relations are a good example. Which university research is of more use for industry: when it is at the direct service of industry interests (no autonomy), when it follows its own path (disconnection), or some compromise between the two extremes? As some studies suggest (Rosenberg and Nelson, 1994), when innovation managers from innovative industries are asked about the most valuable university research for them, they tend to answer that it is the advancement of knowledge in the disciplines more intimately related with their trade. Such answer, in a sense counterintuitive since one would expect that the more research is directly connected to concrete challenges the better, comes from the United States. This is a country where the communication channels between research and industry are extremely diverse and dense. Managers’ answer is understandable: innovation is carried out by creative people, nurtured in research universities where the advancement of knowledge is systematically procured at the highest possible level. We thus suggest that managers’ answer reveals a sort of “connected autonomy” between university and industry. Connected autonomy is a concept borrowed from another context, that of the possible relationships between public universities and society in LA. In the present context, connected autonomy means that, while retaining the freedom to decide the path to be followed by the institution, a good antenna is always screening what is going on in society to better define such path. It is worth recalling that not everything in university research can be defined in tune with current society’s demands. Curiosity-driven research is necessary, not only because serendipity has demonstrated once and again it might open totally new avenues for different kinds of practice, but because good researchers put high value on pursuing an inner logic of search.

59 In what follows, we propose five aspects that STI PM should be thoroughly informed of to design performing policies and we then comment on each of them: i.

Knowledge about the innovative performance of firms and about their absorptive capacities, typically stemming from innovation surveys

ii.

iv.

Knowledge about the overall capacities of the country, for instance through a detailed study of the main characteristics of the NSI or more localized systems of innovation, which in some countries are provided through institutions like STI observatories or national councils of STI Knowledge about the technological needs of the production sectors and other actors, and of diverse public policies that these sectors and actors think would be useful, so that STI policies and instruments can be better tuned with demands from concrete actors Strategic knowledge or foresight on STI

v.

Knowledge about what people think, value and fear about STI.

iii.

The innovative performance of firms This aspect has been empirically explored, with special care, in many countries. The question is if such explorations have been used to inform the design of better policies. Our study and others confirm that the results of innovation surveys have not been a significant input in STI policy-making in LA. This has also been observed in Europe. Arundel (2005: 9) critically remarks: “The CIS (Community Innovation Survey) collects data that could be used to fill some of the gaps in our knowledge of innovation, but unfortunately the CIS has not been fully exploited for this purpose. The main cause is a continued focus on a science-push or linear model of innovation. The countless announcements of the death of this model and its presumed replacement with “systemic” models using Schumpeterian definitions of innovation are definitely premature. The science-push model based on R&D is probably the dominant model in use today by the policy community, although no one refers anymore to it by its name. This has resulted in a lack of demand on the part of policy makers for a wider range of CIS indicators, and a lack of supply from academics and national statistical offices for them”. As an example of wrong policies built on the lack of good indicators rooted on sound theory, Arundel indicates the Lisbon Agenda for the EU, with its aim of 3% of R&D/GDalsoP for 2015, which not only is unattainable but also does not set goals for other parameters of vital importance for innovation. In another study, Arundel (2005) makes a similar observation concerning the divorce between the academic community studying innovation and what PM want to know about how innovation is going on in their countries: “A series of interviews conducted by MERIT staff with members of the European policy community in the Spring of 2005 found that econometric results (stemming from CIS surveys) rarely influenced policy making. Instead, the policy community preferred detailed descriptive analysis, particularly when combined with case studies. This conflicts with the perspective of the academic community, which focuses on econometrics. This has also increased over time, with a decrease in academic reports that contain careful descriptive analyses and a trend towards increasingly complex econometrics in academic publications”. Sometimes PM simply dismiss what academic analyses indicate, in spite of stemming from sound empirical work. For instance, given that providing information services seems reasonable and is easily done, many European and Latin American countries keep on organizing these services, even though innovation surveys clearly state that the lack of information has rarely been scored high among the “obstacles to innovation” options. Why is it that innovation indicators have not been widely used in innovation policy design? In box 3.1, some characteristics of quality and useful innovation indicators are proposed; comments on the difficulty to fulfil these characteristics follow.

60

Box 3.1 – The three qualities of a good innovation indicator (Arundel et al, 1998) The development of effective innovation policy requires good indicators to ensure that our understanding of the problem is correct, good theory to both suggest which indicators are needed and to interpret the resulting data, and an effective policy response to identified problems. Essentially, the goal is to tighten the links between innovation and both government and private actions to improve the innovation process. To help achieve this, innovation indicators should provide information that can meet three requirements: Directly assist the development and implementation of policy actions. The need for indicators to directly assist policy means that the policy significance of each existing and potential indicator needs to be carefully scrutinised. Some indicators could appear to be relevant to policy, when in fact the results could be of little value because political and economic constraints make it highly unlikely that the policy action would ever be implemented. For this reason, the policy value of specific indicators needs to be carefully scrutinised. This requires a good understanding of the policy context, consisting of the existing menu of policy options and the constraints on the potential for developing new policy actions. For example, the current policy context prohibits using tariff barriers to support indigenous new technology firms, although the same goal might be achievable through research subsidies or competitive bidding for government procurement contracts. Verify innovation theory as part of a continual process of testing and improving theories of innovation. The requirement for indicators to improve our understanding of the innovation process is based on the vital role of theory to interpret empirical data. We need indicators that can be used to verify theory and our beliefs and assumptions about the innovation process. An example is the need for indicators to test our theories about national systems of innovation. Assist private firms and other institutions to develop and adjust their own innovation strategies. The social and economic value of innovation indicators will be greatly enhanced if they are of direct value to innovators themselves. For example, indicators that identify best practice can help guide firms and public institutions towards more efficient methods. It is particularly important that indicators obtained from surveys of firms or institutions are of value to them. Managers will be more motivated to complete innovation questionnaires when the results offer clear benefits to their firm. A good indicator should serve as many of these three requirements as possible. This is necessary to keep the questionnaire short while maximising the amount of information that is acquired. The significance of the questions must also be readily understood and lead to direct policy actions. It is of very little help for policy makers if an analysis of indicators comes up with platitudes such as the need to “facilitate the awareness of opportunities and foster the spread of entrepreneurial capabilities”. Instead, innovation indicators (and their analysis) need to provide concrete evidence that can be used to design specific policy actions.

The first characteristic, to directly assist the development and implementation of policy actions, requires clarity and specificity in the policy sphere. If the policy goals are vague, merely quantitative or general (to add knowledge-based value to production, or to increase the proportion of innovative firms in the universe of industrial firms) indicators can be built but they probably will indicate what is already known, for instance that there are relatively few firms that declare having successful innovations introduced in the market. The real challenge is to build good indicators for a delimited and concrete innovation policy. For instance, what do we need to know if we want to help the textile industry to systematically innovate by incorporating high level design? Re-phrasing this question in terms that may be useful for policy design could be: What value do firms attribute to innovating through design? How many textile firms innovate through design? For those that do innovate this way, how do they do it, by subcontracting or in-house? Questions that address directly the information needs of concrete policies can lead to good indicators, but the main difficulty is formulating the right questions. This is particularly critical in developing countries where innovation is important at the discourse level but not always at the practical level of defining concrete concerns that require good tools for innovation policy designs. Uncritical copying of successful policy models is common in LA and these models usually do not lead to raising the right questions so as to inform sound contextualized policies. The second characteristic of a good indicator −verify innovation theory as part of a continual process of testing and improving theories of innovation− is particularly important, and usually missing, in developing countries. There is no universal theory of innovation. Even when a school of thought on innovation has been selected as the most appropriate to orient empirical and analytical efforts, it cannot just be applied everywhere without further refinements and adaptations. Context

61 undoubtedly matters. Let us take for instance the evolutionist approach to innovation theory, that is, the acceptance that (i) firms work in a technological uncertain environment, (ii) some hints about what is going on come from technological natural trajectories, (iii) firms usually work under certain routines within such trajectories and innovations imply challenging such routines. Within this theoretical framework a main question is why firms accept the innovation challenge. There are indicators trying to capture the answer, but there cannot be one universal set of indicators for this. Why firms innovate in a developing country can range from similar reasons than in any country to very specific situations that would rarely been found in a highly industrialized nation. An example of the latter is the request to deliver a product or a service within a very specific and differentiated set of conditions and requirements, leading to a non-standard innovation or to a re-innovation of something already known. If the hypothesis that context matters is taken seriously into account, the set of indicators to answer the question about why firms do innovate in developing countries must be derived from a theory of innovation that recognizes the specificity of such countries, and it will prove its worth by helping to refine such theory. Ultimately, the reason why theory matters is that, without it, processes of trial and error are the only available road to change, and it is too costly. The third characteristic −innovation indicators should be useful for innovators themselves− takes a particular turn when referring to developing countries. Arundel et al (Ibid) example of indicators of best practices can be of relatively little use in developing countries, not only for the overall weakness of innovation practices but because there is a high structural heterogeneity, meaning that best practices will probably be out of reach for the majority of industrial firms. The point is important however, and can lead to original ways of devising indicators. The general question would be: What kind of useful indicators-based analyses can be presented to managers of innovative firms who have answered innovation surveys? Given that innovation is always a collective endeavour involving several actors, signalling to a given firm possible partners in some of the innovative avenues it wants to pursue or reinforce can be useful. This can be made in matrix form, taking key issues signalled by firms and finding who can provide them among actors of the innovation system at national, regional, local or production-chain levels. Unsurprisingly, the application of the Oslo Manual for measuring innovation in developing countries tends to enlarge the gap between research on indicators and the use of its results for policy design. Some efforts have been made in LA to make more sensible the expensive gathering of innovation data, with the Bogota Manual as the main result. However, innovation researchers have usually given priority to comparability over meaningfulness for local conditions, a trend that the Bogota Manual continues to foster. R&D activities, for instance, as defined in the Frascatti Manual, are an important part of LA innovation surveys, in spite of the well known fact that the vast majority of firms in any country of the region do not perform any R&D activity so defined. On the other hand, important information gathered in the OECD countries through the Canberra Manual, namely on human resources devoted to R&D and innovation in firms, is not analyzed in LA, even though some is collected in the innovation surveys. The issue of the knowledgeable people working in firms is important indeed, but few efforts have been made in LA to look at this aspect. In fact, information on how many engineers and of which type (chemical, electrical, mechanical, environmental, etc.) work in firms is neither gathered nor demanded, with few exceptions. This curious fact cannot be attributed to the indifference of innovation theory towards human resources in firms; on the contrary, the importance of human resources has been forcefully highlighted in the literature. Moreover, enquiring about people able to produce innovative solutions in different areas of the firms does not require much theoretical backing. It is of plain common sense that what a firm is able to recognize as important trends depends on the type of workers it employs. Indeed, having either a biotechnologist or a classical agronomical engineer is quite different for the prospects of a seeds producing firm in terms of identifying opportunities and threats for its business. Perhaps this type of information is not so important in countries where the majority of researchers work in business firms, like in Denmark,

62 Finland, France, Japan, Sweden, USA, or South Korea. But this is not the situation of LA. Therefore, identifying the persons able to convey formal and up-to-date knowledge into firms is of great importance for innovation policy design. This is still an unasked question, though. Once the question is asked, it is not difficult to find out an approximate answer and perform some telling analyses: this has been done, for instance, in Uruguay (Bianchi, Gras, Sutz, 2009). What is not so clear is why innovation PM are so little interested in knowing more about something so basic for policy building.39 The overall STI capacities of the country All countries, including developing countries, usually gather indicators on their human resources. UNESCO and the IberoAmerican Science and Technology Indicators Network (RICYT) publish data provided by the S&T governmental offices. The information that is available for almost all countries, at least in LA, consists in educational data, R&D investment by source of funds and by sector of utilization (business firms, government and higher education), and number of researchers by million inhabitants. Other information usually found in OECD statistics is available only for some countries, like distribution of researchers among institutions. The overall picture shows a sharp contrast with the picture that emerges from the average OECD situation: small investment in R&D, with little participation of firms as R&D funds providers or as working places for researchers. Some countries, typically Brazil, have been improving such indicators over the years, reaching more than one percent of R&D/GDP and reporting over 40 percent of researchers working in business firms.40 In any case, the question about the overall capacities of the country should be treated in a way that fosters a better articulation between actors of the NSI, besides indicating the state of S&T indicators. To be used as an articulation tool, indicators must convey both a variety of information and a variety of searching. In Brazil and Colombia, for instance, a complete directory of the research groups is provided on line, with information about the cognitive area and research lines the groups are involved in. In the case of Brazil, the connections of the groups with firms are also indicated. In Uruguay a tool was built with the aim of stimulating articulations between actors of the production and academic sectors, though it will take time until its usefulness can be proved. It consists in a questionnaire, filled by individual researchers and by research groups, where besides describing very succinctly what they are doing, they indicate who could make use of their research capacities. The whole point is to provide systematic information about such capacities that goes beyond anecdotal information. Innovation PM could put this information to good use. An example of this was the policy followed by the Basque Country in the early 1990s, in the midst of its effort to reconstruct the damaged national industrial fabric. The government wanted to modernize the Basque industry and give participation in this effort to the Basque high-tech sectors, particularly microelectronics. It was a clear issue of articulating actors, an issue that required information about the microelectronic firms, their production, and their main clients. The policy makers commissioned this information, and 39

There is another issue adding to this puzzling state of affairs. The fact that LA has tenths times less researchers by thousand inhabitants than the OECD countries is verbalized as a concern and attended at policy level, but the question of where these new cohort of researchers would work is not addressed by PM. Supplyside policies in favour of well trained people are not accompanied by policies to foster the demand of such people in the economy. Perhaps the reason is, again, what Arundel suggested. The “linear model of innovation” is well alive in spite of its repeated funerals, so that working at the very beginning of the chain, claiming that the country has not enough researchers and should get more, appears as all what STI policies ought to do. 40 The data are not totally clear, though, because there is no discrimination between private and public enterprises. In many Latin American countries, and this is so particularly in Brazil, public enterprises are a very important part of industry in strategic sectors like oil and other natural resources, and public institutions that cannot be considered either government or higher education like the agronomic research institutes concentrate many researchers. Thus, the participation of private firms in providing working places for researchers is probably overestimated.

63 provided economic incentives for the effective articulation of firms in need of microelectronics and microelectronics firms able to satisfy firms’ demands. This kind of articulating or networking indicators is of great importance to inform pro-active innovation policies. Organizing and updating them can be a heavy task, but surely one able to put to work together innovation researchers and innovation PM. Knowledge about the technological needs of the productive sectors and other actors Main mismatches between the aims of innovation policies and its results are usually a consequence of the ignorance on the part of PM of what industry really needs to become more innovative. This is not surprising after all, given that what industry needs is not asked in any of the OECD family of surveys. Such surveys ask about what industry does, what is expected from the actions taken, the reasons why some courses of action are no taken, etc., but a thorough inquiry on what industry needs in order to innovate is missing. Perhaps the reason behind this unreasonable state of affairs is that the information needed is expected to come from the combination of what industry wants to achieve plus the obstacles it founds to reach the goals. But this is probably not enough. In innovation surveys the questions around goals and obstacles to innovate are closed questions, with a series of alternatives. In the Latin American case, for instance, such alternatives, provided mainly by the Oslo Manual, leave aside important reasons to innovate and strong obstacles to do it, which is not surprising once it is agreed that innovation is a highly contextual socio-economic process. To give only one example, let us take an obstacle to innovate for a national industrial automation industry that derives from a general policy giving incentives to modernization by subsidizing the imports of automation devices. This sort of obstacles is quite common in developing countries, where the innovation policy is not well articulated with the industrial policy. The latter being usually stronger than the former, enforcing modernization via imports will probably make difficult for a local innovative industrial automation firm to compete. Obstacles are too diverse to be taken into account in closed questions; this is why industrial surveys only allow the analysis of those factors that were presumed beforehand and therefore included as options. Direct and open questions about what industry needs to innovate would be a solution. However, this way of proceeding is not at all common, in part because open questions make surveys more difficult to conduct but also because comparability seems more important than accuracy and meaningfulness. An example of a direct effort to know what industry needs can be found in a recent Argentinean exercise, that was carried out to identify the weaknesses of the production sectors so as to inform innovation and competitiveness policies. The exercise was made in 2008, in the whole country, and it entailed a massive process of consultation. The exercise was jointly undertaken by the Industry Association and the Ministry of S&T. Some of the general conclusions of the study were: The entrepreneurs are clearly interested in a series of transformations that could improve their innovativeness, but implementation requires human resources and other types of support The weaknesses identified are very diverse, depending on the level of development of the sector and its technological complexity In several cases the technological solutions were known, the challenge being its transfer and implementation under local conditions Intersectorial demands were identified, which opens interesting opportunities for industrial articulation and cooperation The connections between firms and the whole system of STI is key for the fulfilment of these opportunities; even more important, there are capacities in the S&T system ready to use for that purpose. These general remarks are illustrated specifically in each of the more than forty sectors that were analyzed in the study. Particularly telling is the following remark: “This type of studies are valuable tools to give visibility to industrial needs and to transform them into demands that can then be jointly addressed in a articulated way with actors of the S&T system”. It is too early to assess the

64 usefulness of this exercise to policy design, but it points to the right direction and also shows that it is possible to gather relevant information by organizing the dialogue with the industrial actors in an open way. It seems clear that researchers following exclusively the set of incentives that rule in academia will not necessarily provide the answers that PM need to design and implement successful innovation policies. It is equally clear that PM who only trust the results for which they pay through consultancy work will not be able to profit from the accumulated research capacities that otherwise may be at their disposal. Some bridges or system of signals must be built to better connect STI research and policy. We propose that the first step be for STI policy to express a clear demand for information and analysis that can put SSH research to work. Examples exist that show that when such demand is expressed the research community responds. This should be then a main focus of future work. Strategic knowledge or foresight on STI Any policy, but innovation policy in particular, needs to face the fact that it is addressing a moving target. Trying to understand where the target will be in a near future is therefore important to design today’s policy interventions. Besides, if a goal is fixed for some years ahead, the road toward it must be designed today. Stating goals for the future implies scrutinizing the present to be able to act. Foresight is, according to Godet (1985), “a reflection for action and against fatality”, characterized by seven key-ideas: i) illuminate present action in the light of the future; ii) explore multiple and uncertain times to come; iii) adopt a global and systemic vision; iv) take into account qualitative factors and actors’ strategies; v) permanently bear in mind that information and prevision are not neutral; vi) choose pluralism and complementary approaches; vii) revise received ideas. There are some key aspects to make foresight useful for policy: i) long term thinking; ii) taking into account the past as well as exploring the future; iii) focusing on a project or a problem; iv) defining an audience (the users of the outcomes are a key factor to take into account when defining foresight exercises); v) engaging different people in participatory exercises; vi) ensuring wide legitimacy to increase the probability that the results will be incorporated into policy design; vii) carefully carrying out foresight because the process can be even more important that the results; and, last but least, viii) having good and reliable information. From all this, it is clear that the STI research community has an important role to play, if the goal is to inform innovation policies. Particularly relevant is preparing the key questions to be answered by a wide variety of specialized people. For example, one of the problems LA should look at very carefully is the prospects of human capital deficits in the Western highly industrialized countries. Such deficit is becoming worrisome for them, under the double pressure of decaying vocation for sciences and engineering (that is being answered through campaigns to enrol women in such vocational directions) and the successful strategies of Easter and Southern Asia to keep on training, retaining and attracting back home their own citizens. If LA, being a quite possible source of recruitment for coping with such deficits, wants to avoid the passive witnessing of a new wave of brain drain, what kind of policy should put forwards? This is a typical policy question that requires the best of foresight to be answered. Knowledge about what people think, value and fear about STI To be successful, any innovation policy needs that whatever is innovative for society −new drugs, new tools, or new procedures− be understood and accepted by society. Innovation policy can act through imposition but this is not a recommended path in democratic societies, both for normative and efficiency reasons. Innovation policies need to communicate with people in some occasions, making explicit the policy goals and the rationale of the selected choices. To be efficient in this communication process, information about what people think, know, value and fear is important. LA and Spain have been doing surveys on the vast issue of public understanding of science, and some of the results are quite informative for innovation policies. But as in the case of innovation surveys, a

65 comparability criterion prevails; there should be more room to detect people perceptions on specific policy topics, for instance, in energy, health, transport, or whatever that will provoke changes in existing routines. This could be a fruitful field for collaboration between innovation PM and researchers, particularly because the process of understanding people’s thinking about STI leads to reflexive analyses about what innovation policies want to achieve: it goes far beyond supporting the innovative behaviour of firms, particularly in developing contexts.

66

Bibliographic references Aghion P, David PA, Foray D (2009), Science, technology and innovation for economic growth: Linking policy research and practice in ‘STIG Systems’, Research Policy 38: 681–693. Ardanaz M, Scartascini C, Tommasi M (2010), Political Institutions, Policy-making, and Economic Policy in Latin America, Working Paper Series, Nr. 158, IDB, Washington. Arundel A (2005), “Innovation Surveys and Policy: Lessons from the CIS”, in Technology Policy Briefs, Volume 4, Issue 1, Maastricht. Arundel A, Smith K, Patel P and Sirilli G (1998), The future of innovation measurement in Europe. Concepts, problems and practical direction, IDEA (Indicators and Data for European Analysis) Paper no. 3. Baptista B (Coord.), Bernheim R, Garcé A, Hernández E (2010), Consulta a tomadores de decisión en políticas públicas de ciencia, tecnología e innovación sobre sus fuentes de información. Informe Regional, Notas Técnicas, No. 154, IDB, Washington. Best A and Holmes B (2010), Systems Thinking, Knowledge and Action: Towards Better Models and Methods, in Evidence & Policy, vol. 6, no. 2: 145-159, The Policy Press. Bianchi, C., Gras, N., Sutz, J. (2009) “Cooperation for innovation as knowledge exchanges: a view from innovation systems”, IEEE Conference Publishing. Booth T (1988), Developing Policy Research, Aldershot: Gower Publishing Company. Cable V (2003), “The Political Context”, Does Evidence Matter Meeting Series, ODI, London. Caplan N (1979), The two-communities theory and knowledge utilization, in American Behavioral Scientist, vol. 22 (3): 459-470. Carden F (2009), Knowledge to Policy: Making the Most of Development Research, IDRC, Ottawa. Cassiolato JE, Lastres H, Szapiro M (2000), “Arranjos e sistema produtivos locais e proposições de políticas de desenvolvimento industrial e tecnológico”. Seminário Local 'Clusters, Innovation Systems and Sustained Competitiveness'. Rio de Janeiro: IE-BNDES, Nota Técnica 5. CHSRF (Canadian Health Services Research Foundation) (2000), Knowledge transfer. Looking beyond health, Toronto: CHSRF. Court J and Young J (2006), Bridging Research and Policy in International Development: an Analytical and Practical Framework, in Development in Practice, Vol. 16, nr. 1: 85-90. Crewe E and Young J (2002), Bridging Research and Policy: Context, Evidence and Links, ODI Working Papers, Nr. 173, ODI. Davies P. (2004), Is Evidence-Based Government Possible?, paper presented at the 4th Annual Campbell Collaboration Colloquium, 19 February, Washington D.C. de Vibe M and Hovland I (2002), Bridging research and policy: an annotated bibliography, ODI Working Paper 174, London: ODI. Delvaux B and Mangez E (2008), Towards a Sociology of the Knowledge-Policy Relation, Literature Revieww Integrative Report. Knowledge and Policy in Education and Health Sectors, Project n° 0288848-2 of the Sixth Framework Program of the EC. Dodgson M, Hughes A, Foster J, Metcalfe JS (2010), Systems thinking, market failure, and the development of innovation policy: The case of Australia, UQ Economics Discussion Paper No. 403, Department of Economics, University of Queensland, and Centre for Business Research Working Paper No.397, University of Cambridge. Dosi G, Llerena P, Labini MS (2006), The relationships between science, technologies and their industrial exploitation: an illustsration through the myths and realities of the so-called ‘European Paradox’, in Research Policy 35: 1450-1464. Erber F (2009), Inovação tecnológica na indústria brasileira no passado recente. Uma resenha da literatura, CEPAL, LC/BRS/R.207, Brasilia. Estebañez ME (2007), Ciencia, tecnología y políticas sociales, in Ciencia, Docencia y Tecnología, No.

67 34, XVIII, May, Universidad Nacional de Entre Ríos European Commission (2008), Scientific evidence for policy-making, Directorate General for Research, Unit L.2-Research in the economic, social sciences and humanities-Prospective, Brussels. European Commission (2007), Towards more knowledge-based policy and practice in education and training, Commission Staff Working Document, Brussels: Commission of the EC. Garrett J and Islam Y ( ), Policy Research And The Policy Process: Do The Twain Ever Meet?, Gatekeeper Series, No. 74, International Institute for Environment and Development. Georghiou L (2007), Demanding Innovation. Lead markets, public procurement and innovation, in Provocation 02: February, London: NESTA. Godet M (1985), Prospective et planification stratégique, Paris: Economica. Grindle M and Thomas J (1990), After the Decision: Implementing Policy Reforms in Developing Countries, World Development, vol. 18 (8). GSRU (Government Social Research Unit) (2007), Analysis for Policy, London: GSRU. Hernes G (2008), The Interface Between Social Research and Policy Making, in European Sociological Review, vol. 24(2):257-265. Hirschman A. (1975), Policy-making and Policy Analysis in Latin America A Return Journey, in Policy Sciences, vol. 6, pp. 385-402. IDRC (2004), Overview: The Influence of IDRC-Supported Research on Public Policy, Ottawa: IDRC. Jones H (2009), Policy-making as discourse: a review of recent knowledge-to-policy literature, ODI Working Papers, No. 5., IKM Emergent Research Programme-ODI, Bonn. Jones N, Datta A and Jones H (2009), Knowledge, policy and power. Six dimensions of the knowledge–development policy interface, ODI Working Paper, No. , London: ODI. Jones N, Jones H and Walsh C (2008), Political Science? Strengthening science–policy dialogue in developing countries, Working Paper 294, ODI: London. Jones N, Pellini A, Pomares J with Datta A (2009), The Role of Evidence-based Policy in Governance Debates: A Synthesis Paper with a Particular Focus on Post-confict Environments, Report commissioned by UNDP Governance Centre, Oslo. Jones N, Pomares J, Pellini A (2009), Think Tanks In Post-Conflict Contexts: Towards EvidenceInformed Governance Reform, Report Commissioned by UNDP Governance Centre, Oslo. Keeley J and Scoones I (1999), Understanding Environmental Policy Processes: A Review, IDS Working Paper No. 89, Brighton: University of Sussex. Keynes JM (1936), The General Theory of Employment, Interest and Money, London: Macmillan (reprinted 2007). Kingdon J W (1984) Agendas, Alternatives, and Public Policies. New York: Harpers Collins. Kirkhart KE (2000), Reconceptualizing evaluation use: an integrated therory of influence. In: Caracelli VJ and Preskill H (ed.), The expanding scope of evaluation use, San Francisco: Jossey-Bass. Knott J and Wildavsky A (1980), If dissemination is the solution, what is the problem?, in Knowledge, Creation, Diffusion, Utilization, Vol. 1(4): 537-578. Lindquist EA (2001), Discerning Policy Influence: Framework for a Strategic Evaluation of IDRCSupported Research, School of Public Administration, University of Victoria. Lomas J (2000), Connecting research and policy, Canadian Journal of Policy Research (ISUMA), No. 1:140-4. Mably P (2006), "Evidence Based Advocacy: NGO Research Capacities and Policy Influence in the Field of International Trade", Working Papers Series on Globalization, Growth and Poverty (GGP), No. 4, Ottawa: IDRC. Marston G and Matts R (2003), Tampering With the Evidence: A Critical Appraisal of Evidence-Based Policy Making, in The Drawing Board: An Australian Review of Public Affairs, Vol. 3, No. 3: 143-163, March, Sydney. Martin BR (2009), The evolution of science policy and innovation studies, TIK Working Papers on Innovation Studies, No. 20080828, University of Oslo.

68 Meek JW (2010), Complexity Theory for Public Administration and Policy, in E:CO, Vol. 12, No. 1: 1-4. Milani C (2009), Evidence-based policy research: critical review of some international programmes on relationships between social science research and policy-making, in Policy Papers, vol. 18. Morlacchi and Martin (2009), Editorial. Emerging challenges for science, technology and innovation policy research: A reflexive overview, in Research Policy 38: 571–582. Mytelka LK (2001), Innovation Theory and Innovation Policy: Bridging the Gap, in OECD, Social Sciences and Innovation, Paris: OECD. Neilson S (2001), “IDRC-Supported Research and its Influence on Public Policy: Knowledge Utilization and Public Policy Processes: A Literature Review”, IDRC-Evaluation Unit, Ottawa: IDRC. Nelson R (1974), "Intellectualizing about the Moon-Ghetto Metaphor: A Study of the Current Malaise of Rational Analysis of Social Problems”, in Policy Sciences, Vol. 5, No. 4, pp.375-414. Nelson R and Winter S. (1982) An Evolutionary Theory of Economic Change, Cambridge, Ma.:Harvard University Press. Nutley S, Walter I, Davies HTO (2007), Using evidence: how research can inform public services, Bristol: The Policy Press. Nutley S, Morton S, Jung T and Boaz A (2010), Evidence and Policy in Six European Countries: Diverse Approaches and Common Challenges, in Evidence & Policy, vol. 6, No. 2: 131-144, The Policy Press. Perkin E. and J. Court (2005), Networks and policy processes in international development: a literature review. Ramalingam B and Jones H with Reba T and Young Y (2008), Exploring the Science of Complexity: Ideas and Implications for Development and Humanitarian Efforts, Working paper 285, ODI: London. Rich A (2005), Think tanks, Public Policy and the Politics of Expertise, Cambridge: Cambridge University Press. Rigby E (2005), Linking Research and Policy on Capitol Hill’, in Evidence and Policy, vol. 1:195-213. Rosenberg N and Nelson R (1994), American universities and technical advance in industry, in Research Policy, 23, 323-348. Sabatier P (1991), Toward Better Theories of the Policy Process, in Political Science and Politics, Vol. 24, No. 2, pp. 147-156, American Political Science Association. Schmidt V and Radaelli C (2004), Policy Change and discourse in Europe: conceptual and methodological issues, in West European Politics, vol. 27 no. 2: 183 – 210. Sfez L (1986), Critica de la Decisión, Mexico: Fondo de Cultura Económica. Stone D, Maxwell S and Keating M (2001) Bridging Research and Policy. An international workshop, funded by DFID, 16–17 July. Warwick University. Sutton R (1999), The policy process: an overview, Working Paper 118, London: ODI. ‘t Velt, R J (2010), “Towards knowledge democracy”, in ‘t Velt, R J (Editor), Knowledge Democracy. Consequences for science, politics and the media, Springer. Texeira F and Ferraro C (2009), “Aglomeraciones productivas locales en Brasil, formación de recursos humanos y resultados de la experiencia CEPAL–SEBRAE”, Serie Desarrollo Productivo 186, Santiago de Chile: CEPAL. UNESCO (2007), Bibliography on the Link(age)s Between Research and Policy, UNESCO, Paris. Weiss C (2003), Studying research utilization, Paper prepared for the IDRC Meeting, Ottawa, March 24-25, Harvard University (unedited) Weiss C (ed.) (1992), Organizations for Policy Analysis: Helping Government Think, London: Sage Publications. Weiss C (1979), The many meanings of research utilization, in Public Administration Review, vol. 39 (5): 426-431. Wingens M (1990), Toward a general utilization theory: A systems theory reformulation of the twocommunities metaphor, in Knowledge, pp. 27-42. 123HelpMe.com (2010), "The Distributive and Knowledge-driven Models of Politics".

69

Annex 1 – Acronyms and abbreviations EU LA LOCTI NSI PM PMP S&T STI

European Union Latin America Ley Orgánica de Ciencia, Tecnología e Innovación National System of Innovation Policymakers Policy-making process Science and technology, or Scientific and Technological Science, technology and innovation

70

Annex 2 – List of interviewed researchers and policymakers Note: the letter following the name of the interviewee indicates wether he/she was interviewed as researcher (R) or policy-maker (PM). ARGENTINA - Mario Albornoz (R) Researcher/Professor, CONICET and Centro Redes Ex-Director of Centro Redes and CAICYT (Centro Argentino de Información Científica y Tecnológica) - Pablo Kreimer (R) Researcher/Professor, CONICET and Universidad Nacional de Quilmes (UNQ) Director, Área de Estudios Sociales de la Cienica y el Conocimiento, Instituto de Estudios Sociales de Ciencia y Tecnología, UNQ. - Gabriela Trupia (PM) Undersecretary, Políticas de Ciencia, Tecnología e Innovación Productiva, Ministerio de Ciencia y Tecnología (MinCyT) - Gabriel Yoguel (R) Researcher/Professor, Instituto de Industria (IDEI), Universidad Nacional de General Sarmiento (UNGS) BRAZIL - Jose Cassiolato (R) Researcher/Professor and Coordinator, RedeSist, Instituto de Economia, Universidade Federal do Rio de Janeiro. - Renato Dagnino (R) Researcher/Professor, Departamento de Política Científica e Tecnológica, Universidade Estadual de Campinas (UNICAMP) - Fabio Erber (R) Researcher/Professor, Universidade Federal do Rio de Janeiro (UFRJ). Ex-General Secretary, Ministério da Ciência e Tecnologia - Priscila Koeller (PM) Adviser to the Executive Secretary, Ministério da Ciência e Tecnologia. - Rafael Oliva (PM) Adviser to the Chairman, Banco Nacional de Desenvolvimento Econômico e Social (BNDES). - Roberto Vermulm (R) Researcher/Professor, Departamento de Economia, Universidade de São Paulo (USP). CHILE - Roberto Alvarez (R) Researcher/Professor, Centro INTELIS de Análisis de la Innovación y Emprendimiento, Departamento de Economía, Universidad de Chile. Banco Central de Chile. - José Miguel Benavente (R)

71 Researcher/Professor, Executive Director, INTELIS, Departamento de Economía , Universidad de Chile. - Carlos Bravo (R) Researcher/Professor, INTELIS Departamento de Economía, Universidad de Chile. - Jorge Katz (R) Researcher/Professor , Facultad de Economía, Universidad de Chile. Ex-Director, División de Desarrollo Productivo y Empresarial, CEPAL. COLOMBIA - Hernán Jaramillo (R) Researcher/Professor and Dean, Facultad de Economía, Universidad Nacional de Rosario. - Mónica Salazar (R) Executive Director, Observatorio Colombiano de Ciencia y Tecnología - Gonzalo Ordoñez (R) Researcher/Professor, Universidad Externado de Colombia Researcher, University of Twente, Netherlands, and Georgia Institute of Technology, USA COSTA RICA - Jeffrey Orozco (R) Researcher/Professor and Research Director, Centro Internacional de Política Económica para el Desarrollo Sostenible (CINPE), Universidad Nacional. - Keynor Ruiz (R) Researcher/Professor, Centro Internacional de Política Económica para el Desarrollo Sostenible (CINPE), Universidad Nacional. CUBA - Jorge Nuñez (R) Researcher/Professor, Director, Estudios de Posgrado, and Director, Cátedra CTS-i, Universidad de La Habana DENMARK - Bengt-Ake Lundvall (R) Researcher/Professor, Department for Business Studies, Aalborg University. NETHERLANDS - Jacqueline Broerse (R) Athena Institute, Vrije Universiteit Amsterdam. MEXICO - Gabriela Dutrenit (R) Researcher/Professor, Programa de Posgrado en Economía y Gestión de la Innovación, Universidad Autónoma Metropolitana (UAM-Xochimilco). - José Luis Fernández (PM) Researcher/Professor, Instituto de Ingeniería, Universidad Nacional Autónoma de México (UNAM) Ex-General Coordinator, Foro Consultivo Científico y Tecnológico. - Lorenza Martinez (PM) Undersecretary, Industria y Comercio, Secretaría de Economía.

72 - Luis Mier y Terán (PM) Director, Sistema Nacional de Investigadores CONACYT - Martín Puchet (R) Researcher/Professor, Facultad de Economía, Universidad Nacional Autónoma de México (UNAM). - Leonardo Rios (PM) Deputy Director, Desarrollo tecnológico y Negocios de Innovación, Consejo Nacional de Ciencia y Tecnología (CONACYT) - Leopoldo Rodriguez (PM) Member of Board of Directors, Asociación Mexicana de Directivos de la Investigación Aplicada y el Desarrollo Tecnológico (ADIAT) - Daniel Villavicencio (R) Researcher/Professor, Universidad Autónoma Metropolitana (UAM-Xochimilco) SOUTH AFRICA - Jo Lorentzen (R) Human Sciences Research Council, Cape Town. - Jyshree Naidoo (PM) Manager, Innovation &Entrepreneurship, Development Bank of South Africa. SPAIN - Elena Castro (R) Researcher/Professor, INGENIO (CSIC-UPV). - Ignacio Fernandez de Lucio (R) Researcher/Professor, INGENIO (CSIC-UPV). SWEDEN - Claes Brundenius (R) Honorary Professor, Research Policy Institute, Lund University, Sweden UNITED KINGDOM - Joanna Chateway (R) Director, Innovation and Technology Policy, RAND Europe Co-Director, ESRC Innogen Research Centre, Development Policy and Practice, Open University. URUGUAY - Gerardo Agresta (PM) Director, Dirección de Innovación, Ciencia y Tecnología para el Desarrollo (DICYT), Ministerio de Educación y Cultura (MEC) - Pablo Alvarez (PM) Member of Parliament. President, Comisión de Ciencia y Tecnología, Cámara de Senadores. - Luis Bértola (R) Researcher/Professor and ex-Dean, Facultad de Ciencias Sociales, Universidad de la República (UDELAR). - Rafael Canetti (PM) Member of Board of Directors, Agencia Nacional para la Investigación y la Innovación (ANII)

73 Researcher Professor, Facultad Ingeniería, Universidad de la República (UDELAR). - Lucía Pittaluga (R) Consultant, Unidad de Políticas y Programas, UNDP-Montevideo Researcher/Professor, Cátedra de Crecimiento y Desarrollo, Facultad de Ciencias Económicas y de Administración, Universidad de la República (UDELAR). - Edgardo Rubianes (PM) President, Board of Directors, Agencia Nacional para la Investigación y la Innovación (ANII) Researcher/Professor, Facultad de Veterinaria, Universidad de la República (UDELAR). - Judith Sutz (R) Researcher/Professor and Coordinator, Unidad Académica, Comisión Sectorial de Investigación científica (CSIC), Universidad de la República (UDELAR). USA - Susan Cozzens (R) Associate Dean for Research, Ivan Allen College Director, Technology Policy and Assessment Center Georgia Institute of Technology. - Calestous Juma (R) Professor of the Practice of International Development and Director of the Science, Technology and Globalization Project, Harvard Kennedy School, Harvard University. - Richard R. Nelson (R) Professor Emeritus, Columbia University. VENEZUELA - Maria Antonia Cervilla (R) Researcher, Universidad Simón Bolivar - Luis Marcano (PM) Asesor de organismos públicos Ex-Viceminister, Planificación y Desarrollo, Ministerio de Ciencia y Tecnologia - Alexis Mercado (PM) President, Centro Nacional de Tecnología Química (CNTQ), Ministerio del Poder Popular para Ciencia, Tecnología e Industrias Intermedias. - Nuris Orihuela (PM) President, Agencia Bolivariana para Actividades Espaciales (ABAE) Professor, Universidad Central de Venezuela (UCV) Ex-Minister, Ministerio de Ciencia y Tecnología - Rafael Palacios (R) General Director, Dirección de Gestión de Innovación, Instituto de Estudios Avanzados (IDEAS) - Arnoldo Pirela (R) General Coordinator, Laboratorio de Innovación y Aprendizaje (LIA), Researcher, Centro de Estudios del Desarrollo (CENDES), UCV . - Nydia Ruiz (R) Coordinator, Proyecto Gestión de Conocimiento, Vicerrectorado Académico, UCV - Hebe Vessuri (R)

74 Director, Departamento de Estudios de la Ciencia, Instituto Venezolano de Investigaciones Científicas (IVIC).

75

Annex 3 – Questionnaires used for the survey

Analysis of the linkage between social sciences and humanity research on innovation and STI policy development

QUESTIONNAIRE FOR INTERVIEWS TO RESEARCHERS WHO PARTICIPATE IN SOCIAL SCIENCES AND HUMANITY RESEARCH GROUPS ON INNOVATION (2010)

Note to interviewees 1. This questionnaire is being addressed to different research groups in Latin America and Europe. 2. It aims at obtaining information essentially about:  the research group to which you belong, its interests and achievements in fields related to innovation policies;  your perception regarding the design and implementation process of innovation policies in your country;  the relationships between your research group and the decision makers in the field of science, technology and innovation (STI) policies. 3. Your answers will be kept confidential. They will be analysed exclusively by the CSIC team assigned to EULAKS. Your name will be included in the list of interviewees in an annex but it will never be cited in the study. 4. The following abbreviations are used in the questionnaire:  STI: science, technology and innovation  Group: innovation research group in social sciences and humanity

1.1. Name 1.2. Position and institution where you work as a researcher 1.3. For how long have you been working on innovation related issues? And what was the trigger to this orientation in your career? 1.4. Education Module 1.A 1.5. Name of the research Group you belong to and main participants (or Web page on the Group) 1.6. What are the main interests guiding the Group's research (beyond its formal objectives) and on which types of actors does research focus? 1.7. How does the group identify and select its research projects? (Are there some specific mechanisms?; do they sometimes respond to direct demands from the government or other specific actors?) 1.8. Does the Group intend or pretend influencing STI policy, either on design, implementation or evaluation issues?

No (explain) Partially Much It is a specific objective

If this is so: 1.8.1. At which of the following levels?

National STI policy Regional STI policies Sectorial STI policies Crosscutting issue (which one?) Some specific issue (which one?)

1.8.2. How do you seek to influence? (Other diffusion mechanisms than academic papers? Promoting public debates? Through intermediary organizations between knowledge production and use? Other ways?) 1.9. Could you indicate 3 to 5 authors whose work has been particularly useful to guide the Group's research?

77

1.10. Has the Group generated any concept that was later taken up either in the sphere of policy-making or research (national or international)? 1.11. If you would have to describe your Group with 5 descriptors, which ones would you choose? (excluding science, technology and innovation) 1.12. Do you consider that one or several studies carried out by the Group have become at some point a reference in innovation related discussions at the national level or abroad? If this is the case, could you identify them? 1.13. Can you identify any study of your Group that you consider should have been taken into account –but has not–, in the sphere of STI policy design? Why do you think this happened? 1.14. Do you know other research groups or individual researchers in your country that study innovation from a social sciences or humanity perspective? (please name some of them) In this case: 1.14.1. What would you say is the main difference between your Group and these groups or persons, in terms of research? 1.14.2. Do all these groups interact? 1.15. What type of relationship do you have with research groups of other countries? 1.16. Do you know individual researchers who work on innovation without forming part of a group on innovation in your country? Do you know any of their publications and do you interact with them? Modules 1.B y 1.C 1.17. Would you say that this instrument and the other existing ones are articulated into a national STI or innovation policy?

Author, title of the study, year

78

If this is (partially) so: 1.18. Do you consider that this STI policy is in tune with the approach to innovation of your Group? If this is (partially) not so: 1.19. What do you think is missing or needed in your country in order to have an innovation policy or, more generally, a STI policy? 1.20. As far as you know, what are –in your country– the main inputs and factors taken into account in the definition of the STI policy or –if there isn't any policy– of promotional instruments? Please, intend to identify 5 inputs and try to prioritize them. Take into account that there is an item "others" and that the influence can be either direct or indirect.

Publications of scientific or technological research outcomes (any science) Outcomes of deliberations between researchers and policy-makers Outcomes of action projects carried out by the government or NGO Quantitative data obtained through surveys or similar Analysis or conclusions from committees on specific issues Studies and diagnostics of economic sectors, and their policy lessons Prospective studies Personal knowledge and experience of the policy-makers Personal or political interests of policy-makers Pressures from advocacy groups, interest groups or lobbies Working lines and financing of international organisations Budget negotiations at the national level (resources assigned to STI development and promotion) Public opinion Others (specify):

1.21. Who are the main actors who presently define the STI policy and/or its promotional instruments? (We refer to influential direct and indirect actors) 1.22. How would you describe the relations between these actors and your research Group? 1.23. Finally, what do you consider are the main obstacles and difficulties for a better linkage between social sciences research on innovation processes and decision making in the field of STI policy and instruments? How do you think this linkage could be improved?

79

Analysis of the linkage between social sciences and humanity research on innovation and STI policy development

QUESTIONNAIRE FOR ACTORS INVOLVED IN THE DESIGN OF INNOVATION POLICY

(2010)

Note to interviewees 5. This questionnaire is applied to policymakers in different Latin American countries and some Europeans. 6. Through this interview we are looking for information essentially related to:  the process of innovation policy development in your country;  your perception of the relevance of local research groups in social sciences and humanities (SSH) on topics related to the innovation process;  the difficulties to bridging SSH research and policy-making in the field of science, technology and innovation (STI). 7. Your answers will be kept confidential. They will be analysed exclusively by the CSIC team assigned to EULAKS. Your name could be included in the list of interviewees in an annex but it will never be cited in the study.

2.1. Name of the interviewee 2.2. Position and institution from where you participate as a policymaker 2.3. For how long have you been in this position and what was your previous position? 2.4. Education Module 2.A 2.5. ¿Would you say that a STI or an innovation policy presently exists in your country? In case it exists: 2.6. ¿Could you briefly explain how this policy and its instruments have been defined, and who are the key actors in this process? 2.7. ¿What is the relation of this policy [or, if there is no policy, of the innovation promotional instruments] with the economic and social development policies of the country? 2.8. ¿Do you consider that the needs and demands for innovation have been reasonably identified in your country, in the different fields that should orient an innovation policy? 2.9. ¿Should something been done in this regard? Publications of scientific or technological research outcomes (any 2.10. ¿From the following science) list, what do you consider are the main inputs taken into account in the design of the innovation policy or promotional instruments in your country?

Outcomes of deliberations between researchers and policy-makers

Please, take into account that:

Sectoral studies and diagnostics, and their policy lessons

- there is an option "others"; - inputs can be direct or indirect; and - we would like you to rank the options you select.

Outcomes of action projects carried out by the government or NGO Quantitative data obtained through surveys or similar Analysis or conclusions from committees on specific issues

Prospective studies Personal knowledge and experience of the policy-makers Personal or political interests of policy-makers Pressures from advocacy groups, interest groups or lobbies Working lines and financing of international organisations Budget negotiations at the national level (assignation of STI resources) Public opinion (eventually channelled through an organization or other means of communication) Others (specify)

81 Module 2.B y 2.C 2.11. ¿Do you know research groups or individual researchers in your country, who study issues related to the innovation process from a social sciences or humanity (SSH) perspective? Please name them. 2.12. ¿Can you remember if some studies of these researchers have been considered as references for the development of innovation policies, programs or instruments?

Authors, titles and approximate publication year:

2.13. ¿In general, what do you think should be the role of SSH research with regard to the definition of innovation policies in the country? 2.14. Looking at the facts in your country, do you consider that this type of research:

Please, explicit your answers.

2.15. The international literature often refers to the non-linearity of the research-policy linkage, and to the decisive role of mediators in this process. Can you mention some organizations or groups who play this function in your country? 2.16. Summing up, what would you say are the main obstacles or difficulties to link SSH research and policy-making in the field of STI? 2.17. ¿How do you think it would be possible to improve this bridging and, in general, the use of SSH for innovation policy purpose in your country? 2.18. Lastly, are the impacts of promotional instruments being evaluated? If some are, who conducts the evaluations? Is research involved?

should address other problems than the ones presently addressed? should be carried out in other ways than the present ones? should be diffused or communicated in other ways than the prevailing ones?