Research-Policy Relations and Migration Studies

  • Peter ScholtenEmail author
Open Access
Part of the IMISCOE Research Series book series (IMIS)


Migration research and policymaking can interact in many different ways. Sometimes policy actors are directly or indirectly involved in defining migration research, for instance, by commissioning specific studies or developing funding schemes with specific objectives. Alternatively, research can also play a role in agenda setting, defining policy problems and triggering policy responses. This chapter examines how different configurations of research-policy relations can affect (and have affected) qualitative (as well as quantitative) migration research. The chapter looks at policy-related as well as academic factors that play a role in configuring mutual relations, at the European and the national level as well as the local level. Amongst others, the chapter will discuss the role of European and national funding schemes, advisory bodies, knowledge institutes and the role of international research networks such as IMISCOE. In particular, the chapter makes the argument that research-policy relations, when involving strong relations of mutual interdependency, could have a constraining effect on research questions and methodologies in migration research. This applies to qualitative as well as quantitative research. A more reflective attitude in research-policy dialogues could, in contrast, lead to more critical reflection based on migration research and prevent paradigmatic closure.

16.1 Introduction

Migration research has evolved rapidly as an interdisciplinary research field over the last three to four decades. Until the 1980s, studies of migration were relatively scarce and generally fed into generic disciplines as in particular sociology and anthropology, rather than in the conceptualization and theorization of migration per se. Take, for instance, Thomas and Zaniecki’s (1926) study on Polish migrant peasants in the United States that contributed to a broader functionalist perspective on societal assimilation. Especially since the 1980s, more and more scholars from a broader range of disciplines developed an interest to migration, and directly related also to the process of migrant incorporation (see also Chap.  4, this volume). Nowadays, migration research is a broad international and strongly institutionalised research field with scholars from various disciplines (Brettell and Hollifield 1994; Bommes and Morawska 2005).

The evolution of migration research as a research field, defined here more broadly as a field of study on migration itself as well as migration-related diversity and immigrant integration, has been shaped by many factors. Besides the growing prominence of mobility and diversity as a social fact, the broader social and political environment of this research field has played a particular role. On the one hand, the political sense of urgency around migration and diversity has provided many opportunities for migration research to have an impact on policymaking and on societal discourse more in general. Take, for instance, the many expert-led government committees and knowledge brokers that have laid the foundation of migration polices in many countries. Even today, migration researchers are a prominent voice in the public debate around issues like the refugee crisis or radicalization, and a broad range of institutes has evolved operating at the boundaries of science and politics to contribute to ‘evidence based policymaking.’

On the other hand, the political and broader social setting has also influenced the development of migration research itself. Developments in the world around us, including trends and events as well as for instance funding structures, can influence how we do migration research. Some even speak of ‘coproduction’ of knowledge, which means that the policy environment exerts influence on the types of knowledge paradigms that emerge, or perhaps those that do not emerge (Entzinger and Scholten 2015).

In terms of qualitative research, this mutual relationship can impact research in direct as well as indirect ways: it can influence the type of questions migration scholars ask (and perhaps do not ask), the concepts and theories they use (or disregard) and how they relate to policymakers in doing research and in disseminating research findings. For instance, Favell (2003) has argued that research-policy relations in the past sustained a so-called ‘integration paradigm’ that framed academic discourses in terms of ‘integration’, while at the same time legitimizing government interventions in terms of integration. The mutual relationship between research and policy will affect quantitative research as well, as for instance visible in the very lively debates on how and why to collect data on factors such as ethnicity (Simon et al. 2015).

This chapter provides a conceptual and analytical framework for making sense of research-policy relations and the impact on both research and policy. First, an overview is provided of different ways of configuring the research-policy nexus. Distinguishing various ideal-types helps making sense of the diversity in research-policy relations that one can encounter in practice. Secondly, knowledge production is discussed, with a particular emphasis on how research-policy relations might have influenced knowledge production in migration research over the past decades. Thirdly, the attention shifts from production to knowledge utilization. This includes distinguishing between various types of knowledge utilization (Fig. 16.1).
Fig. 16.1

The three main aspects of research-policy relations (Scholten et al. 2015)

Finally, the chapter will explore implications of these insights for how to engage in research-policy dialogues as a migration scholar. What lessons can be drawn from studies in the abovementioned framework; how to contribute to reflexivity on behalf of migration scholars when engaging in research-policy relations? Here also a connection will be made to sociology of science literature which provides important insights on how scholars are also actors in the definition of different types of research-policy relations, knowledge production and utilization.

16.2 Configurations of the Research-Policy Nexus

There are very different ways of configuring relations between social science research and politics, both on a normative and an empirical level. Often, debates on research-policy relations are highly normative and mobilise specific models or ‘discourses’ of how research-policy relations should be configured. Policymakers sometimes complain that the research they receive is insufficiently instrumental, thereby discursively mobilizing a model in which social science research should be instrumental to policymaking. Or researchers mobilise a model in which they enlighten policymakers (“speak truth to power”) as real academics, whereas in practice their research may often be ignored or perhaps only very selectively used.

One discourse that has obtained particular prominence is ‘evidence based policymaking’ (Sanderson 2002). This suggests that policymakers make use of evidence from research when designing or implementing policies or in decision making. The concept is, however, rather abstract as it says little about whether it was the evidence that confronted and eventually made policymakers take specific actions, or whether it was the policymakers who actively searched for those knowledge claims that were most convenient to their plans. It also says little about how this evidence was used in policymaking, whether it was used in a direct and instrumental way or for instance in a more symbolic way, and whether it involved actual relations between researchers and policymakers or whether policymakers had other ways to obtain evidence for policymaking.

Hence there is a need for a clearer framework to study how research-policy relations actually are configured in empirical practice. In this context, a distinction is made between various ideal-type research-policy relations that differ on two dimensions (Scholten 2011; Hoppe 2005; Wittrock 1991). First, whether either research or politics has primacy in mutual relations. Is it research that puts new issues on the agenda or gets to develop innovative policy solutions, or is it politics that gets to select those knowledge claims that it sees fit and ignore others? Secondly, whether there is a sharp distinction between the roles of policymakers and researchers or whether their roles are more or less entwined. For instance, to what extent are scholars actively engaged in policy processes, or do scholars rather stay in the so-called ‘ivory tower’ protected by scientific objectivity?

Four ideal types are often distinguished in the literature. First, the enlightenment model, which is perhaps the most classical model of how scientists see their relationship to other realms such as politics. In the enlightenment model, the academic does not directly engage in policy processes, but contributes to the development of public knowledge and to defining policy problems via traditional academic means (such as books, articles). There is a clear differentiation in roles and academics assume a role of primacy in mutual relations; they are conceptual and theoretical ‘brainwashers’, whose concepts and knowledge gradually creeps into society and determines how policymakers act (or do not act). Enlightenment can take place through quantitative as well as qualitative research, but specific to qualitative research can be the impact of key concepts that are developed in social science research. Think, for instance, about how key concepts as assimilation or social capital have framed public understanding of migration and diversity, or how key readings such as Putnam’s “Bowling Alone” has been read so widely that it becomes part of common public understanding of communities and diversity.

In the technocratic model, research also assumes primacy in research-policy relations, but does so in a much more direct way. Technocracy means that the roles of research and policymaking are hardly distinguished, and that scientists can, to some extent, take over the role of policymakers and politicians. They do not only produce knowledge, but are also involved directly in the production of policies. Often this type of research-policy relations is associated with depoliticisation. Take, for instance, the many cases where expert committees laid the foundation of government policies. For instance, expert committees such as the Cantle committee led the foundation of the UK’s community cohesion policies, or in the Netherlands the Scientific Council for Government Policy provided a direct impetus to key policy turning points at several moments over the past decades.

Thirdly, in the bureaucratic model, it is rather politics that is on top and research that is on tap. In this model, the roles of research and policymaking are sharply differentiated, with research producing ‘facts’ and policymaking taking normative decisions based on the facts. This ‘fact-value’ dichotomy creates a sharp boundary between both worlds; scientists are not supposed to engage in any value laden debate and therefore stay far from actual political decision making. In many countries, there are statistics offices that produce data in relation to migration and diversity; research (Simon et al. 2015) shows that often these offices produce data in a way that is in accordance with the particular social and political environment in which they operate. One example is the taboo on ethnic statistics in France in contrast to the custom of producing ethnic statistics in the UK and in the Netherlands. Think also about (qualitative) policy evaluations, for instance using the multiple-stakeholder analysis approach to find out whether actors perceive a specific policy intervention as successful, rather than in itself attributing new idea for alternative interventions.

Finally, in the engineering model, the roles of research and politics are once again more entwined, but politics preserves clear primacy in mutual relations. This will often mean that politicians and policymakers actively ‘pick and choose’ those scholars and those knowledge claims that they see fit for a particular purpose. For instance, if politicians have already decided to pursue a certain course of policy action, they may search for research that can help to substantiate this policy action and to make sure that it is properly implemented. This type of boundary configuration can also involve the deployment of research as a form of ‘political ammunition.’ This can involve qualitative as well as quantitative research, depending on the type of ammunition that is required in a specific setting. In qualitative research, comparative methods are often used to this aim, to find out what works in what section and to transfer clear policy lessons (policy diffusion, policy transfer) wherever possible (Table 16.1).
Table 16.1

Theoretical models of research-politics relations


Coordination or relations

Scientific primacy

Political primacy

Demarcation of roles


Enlightenment model

Bureaucratic model


Technocratic model

Engineering model

Adapted from Scholten (2011), Hoppe (2005), and Wittrock (1991)

These ideal types help to develop a better understanding of the diverse types of research-policy configurations one may encounter in practice. They are ideal types, which means that in practice one is more likely to encounter blends of different types, or ‘in between’ types. Take, for instance, the fact-value distinction in the bureaucratic type, which will never be ‘pure’ in empirical practice; the choice of what type of facts to produce and which facts to ignore is in itself a normative decision (see also the normative debate on the role of ethnic statistics in migration research). There may be differences between countries in terms of the models of research-policy relations that have developed and institutionalised historically. In some countries, such as France, with a very state-centric and politicised policy process, there may be a stronger political primacy in mutual relations, often leading to either a bureaucratic or engineering configuration. Other countries, such as the Netherlands and, to some extent, also the UK, have a strong tradition in directly engaging research in policymaking (technocracy). However, in practice, various models will often coexist in particular policy settings.

The various ideal types can also involve different configurations of qualitative and quantitative research. For instance, much work has been done on how ethnic statistics helped sustain government interventions via a bureaucratic model, providing the ‘facts’ so that to sustain specific political ‘values’. But qualitative research can apply to all four configurations as well. Think of how phenomenological or ethnographic research can contribute to unique new insights and thus contribute to enlightenment. Or how qualitative research can help develop policy interventions as in the technocratic model, which, according to Favell, has been the case in the context of the so-called integration paradigm. Or qualitative research such as multiple-stakeholder analysis that provide the facts for more normative evaluations of policies. Or qualitative approaches such as social action research that can provide ammunition for or against specific government interventions.

Furthermore, developments in the broader social and political context can lead to important changes in the type of boundary configuration that emerges. One such development that has manifested itself throughout Europe over the past decade or so is the growth of political contestation, or politicization, of migration and diversity. Research shows that politicization does not lead to a deconstruction of research-policy relations, but rather to a reconfiguration of types of relations (Scholten and Verbeek 2015). In particular, it leads to the emergence of the two types with relative political primacy, the bureaucratic and engineering models. At the same time, relations that match the enlightenment or technocratic models met with increasing contestation; a process of politicization of research took place in which the credibility of scholars was frequently put on the line. In the Netherlands and the UK, this led to frequent accusations against scholars who had been involved in policy developments in the past, for having been ‘biased’ in favour of the now much despised multicultural policy model (Scholten 2011). Caponio et al. (2015) show that this also led to competition within the academic world between scholars from various research paradigms that now increasingly had to compete for attention in the politicised setting.

A particular role in the constitution of specific types of research-policy relations is played by organisations that operate ‘in between’ or ‘on the boundaries’ of research and politics. In the literature, these are also described as ‘boundary organisations’ (Miller 2001). In the field of migration research, many different boundary organisations exist at the European and the national level as well as (on a more limited scale) at the regional and local levels. Notable examples in the field of migration and diversity policies are the Migration Policy Institute, the Migration Policy Group, the Migration Policy Center and the Center for European Policy Studies. These boundary organisations can operate in terms of various of the models discussed above, such as enlightenment or engineering ones. Another specific type of boundary organisations involves ‘independent commissions’ that are often installed on an ad-hoc basis and for a limited duration in response to events or developments that induce governments to gather new knowledge and information (Boswell and Hunter 2015). Especially the UK has developed a tradition to install such independent commissions after events, such as ethnic riots in UK cities in 2001 and the terrorist attacks in London in 2005 (ibid).

In sum, the type of research-policy nexus may also have influence on developments in both the fields of policymaking and that of research itself. It may affect patterns of knowledge utilization, for instance with promoting direct and instrumental forms of knowledge utilization in the technocratic model, whereas the engineering type would promote more symbolic and indirect forms of knowledge utilization. And it may affect knowledge production in the field of research, for instance, by privileging specific research actors, knowledge claims or institutes, or in contrast by ignoring others. This can affect the type of questions that scientists ask, the type of knowledge paradigms that emerge and the type of methods that are used. In the following two paragraphs, these two forms of impact will be discussed more in depth.

16.3 Knowledge Production

A key lesson to be learnt from the sociology of sciences and the sociology of knowledge is that social scientific knowledge is not produced in a social vacuum (Gieryn 1999). Some speak of ‘co-evolution’ or ‘co-production’ of knowledge in interaction between social scientists and their broader social and political environment (Jasanoff 2005). The opportunity structures that a policy context offers in terms of influencing policy making will in turn also influence the field of migration research in terms of methodological, theoretical and disciplinary developments.

In terms of qualitative research, there are many examples of how research-policy relations have impacted migration research. Speaking of the notion of coproduction of social scientific knowledge, various migration scholars have pointed at the key role of the nation-state as a ‘constitutive frame’ for the development of migration research, especially in the 1980s and 1990s (Thränhardt and Bommes 2010). This would have promoted a national ‘container view’ without much regard for similar processes in other settings. Furthermore, it would have induced an orientation on contributing to problem solving rather than on theoretical development of migration research as such. Favell (2003) captures this in terms of the coproduction of an Integration Paradigm in migration research, between, on the one hand, policymakers interested in instrumental knowledge for promoting integration and, on the other hand, researchers with a strong policy – rather than theoretical orientation.

This account of the national coproduction of the Integration Paradigm is particularly illustrative for how the concepts and theories that migration researchers use develop in specific social and political settings, and will also carry tacit assumptions from these settings. In this case, one of such tacit assumptions is that there is an instrumental need for integration of newcomers within nation-states. Wimmer and Glick-Schiller (2002) have taken this argument even further by arguing that this coproduction has led to a tendency of ‘methodological nationalism’ in migration research. According to them:

nation state building processes have fundamentally shaped the ways immigration has been perceived and received. These perceptions have in turn influenced, though not completely determined, social science theory and methodology and, more specifically, its discourse on immigration and integration (ibid, pp. 301–302).

Indeed, various scholars have emphasised how migration research often tended to reify specific national models of integration (Bertossi 2011; Scholten 2011; Bertossi and Duyvendak 2012). A national model would involve a nationally and historically rooted approach to migrant integration, which would be strongly institutionalised in national policies and would determine the national discourses on migrant integration. Examples include the French Republicanist model of integration, the British race relations model, the American ‘salad bowl’, or the multicultural model that would have typified Dutch policies for a long time. These national models are also manifested amongst others in the labelling of migrants or in the ways of data collection in those countries. For instance, in the UK migrant communities were often framed as ‘racial minorities’, in the Netherlands as ‘ethnic minorities’, in Germany as ‘foreigners’ and in France there was a taboo on labelling migrants as that would conflict with the colour-blind Republicanist model.

Similarly to the coproduction of knowledge at the national level, there is evidence of a similar process of coproduction, but then between European institutions and migration scholars. Geddes (2005) speaks of the coproduction of migration as a European problem, where migration scholars help problematise migration (and diversity) in such a way that a European response would be necessary. Geddes and Scholten (2015) show how this relationship between the EU and migration research is also substantiated by various funding schemes (such as the European Integration Fund, and the European Asylum and Migration Fund). These funds are often designed to bring together scholars to work on issues of relevance to the EU’s agenda. In turn, this has promoted a focus amongst migration scholars not only on comparative research, but also on those questions and those issues that are of particular relevance to Europe.

This coproduction is, of course, not limited to qualitative migration research. In fact, the production of statistics on migrants (or ‘ethnic statistics’) has always taken in a central position in the discussion on coproduction of migration research. Some countries, such as the UK and the Netherlands, have long traditions in collecting so-called ‘ethnic statistics’, or statistics that monitor the social, economic and cultural position of migrant groups or ‘minorities.’ How these countries collect data reveals remarkable differences. These are related, for instance, to what categories are applied (minorities, migrants, foreign-born, race) as well as how the data is collected (public census, classification based on official populations statistics). Some countries, such as France, have always fiercely opposed monitoring based on ethnicity or race. These consider ‘ethnic statistics’ as opposed to the colour-blind orientation that may be expected from governments, or consider ethnic statistics as social constructions that may have a performative or labelling effect that leads to the exclusion rather than inclusion of minorities.

The collection of ethnic statistics has been object of fierce debate within academia as well as in politics. A key criticism from academia is that by collecting data based on ethnicity (or culture or race), researchers inadvertently legitimise government intervention based on these categories (Favell 2003). Collecting numbers on specific ethnic groups leads to a reification of the image that these ethnic groups really exist. Subsequently, the attribution of ‘problems’ to those groups, for instance by showing that school drop-out rates are higher amongst specific ethnic groups, legitimises government intervention (Rath 2001). Such government intervention may then again label or even stigmatise those groups as ‘problematic’ also in public and political discourses. Also, many sociologists have objected that ethnicity, race and culture cannot be seen as clear markers of groups, as they involve social constructions (what does ethnicity really mean?) and often reveal remarkable internal diversity within these groups (the Asian populations involve many very different groups with different beliefs, cultures and social and economic backgrounds).

However, others have contended that the absence of data on the position of migrants and minorities may also legitimise ignorance towards the problems that migrants often face. For instance, Amiraux and Simon (2006) argue that the absence of ethnic data in France has legitimised a ‘non-policy’ towards migrants. Amongst others, the absence of ethnic data makes it hard to address discrimination problems, especially institutional discrimination or racism which can only be identified with help of relevant data. In the Netherlands, ethnic statistics have also, for a long time, been used to finance primary schools, with schools receiving more funds for every child from an ethnic minority group.

Efforts have been made to overcome the methodological nationalism in the collection of ethnic statistics, and design a more comparative approach. European projects such as PROMINSTAT and COMPSTAT have been developed in close cooperation with the EU to do precisely that. Also, the MIPEX (Migrant Integration Policy Index) has been developed in collaboration with the EU, to get more comparable indicators on migrant integration policies. However, here too an element of coproduction is manifest, with the EU particularly interested in promoting some form of policy convergence by means of facilitating mutual learning, in an area where the EU does not formally have strong competencies.

A recent review of the development of migration research as a research field (Scholten et al. 2015) shows that especially over the last decade or so, migration research has strongly internationalised. This evolution of international research networks such as IMISCOE (International Migration, Integration and Social Cohesion in Europe) and the growing interest in more comparative research (Bloemraad 2013; Saharso and Scholten 2013) have played a key role in this regard. Steered by such international comparative research, the field of migration research has increasingly developed its own theoretical orientation as a research field, with its own body of literature, network events and even training programs such as PhD training facilities and various international masters programs throughout Europe.

16.4 Knowledge Utilization

A third aspect of research-policy relations that is discussed in this chapter involves knowledge utilization. Here, three basic questions can be asked; knowledge utilization in what way, of what, by whom and when? Concerning the question of how knowledge is utilised, Christina Boswell (2009) distinguished between two types of knowledge utilization: instrumental and symbolic knowledge utilization. Instrumental knowledge utilization means that knowledge is used directly in policymaking for instrumental purposes, such as designing better policy alternatives, taking a better decision, making sure that policies are implemented and to evaluate past policies. This type of knowledge utilization speaks most closely to the notion of ‘evidence based policymaking’. Symbolic knowledge utilization involves a type of knowledge utilization in which knowledge is not used directly for the development of new or better policies, but indirectly for other purposes related to the policy process. Boswell differentiates two types of symbolic knowledge utilization. First, substantiating knowledge utilization means that knowledge is used primarily to substantiate specific policy claims or ideas. This means that knowledge and expertise is selected from the available stock in order to lend support, credibility or authority to an already existing policy claim. A particular kind is to use research as a form of political ammunition in political debates and conflicts. Secondly, legitimising knowledge utilization does not so much lend support to a specific claim, but to a specific actor involved in the policy process. For instance, actors can mobilise knowledge and expertise as a mere symbolic act to claim authority in a field or to simply illustrate to others that they are taking a policy problem very serious and hence mobilise knowledge. The mere existence of government-associated research bodies on migration and integration, such as the BAMF in Germany, helps to legitimise the role of (specific parts of) governments in intervening in migration and integration. A particular form of legitimizing knowledge utilization involves the so-called ‘fridge’ function of research. When confronted with a wicked or contested policy problem, policymakers and politicians may be tempted to temporarily remove the problem from the agenda by announcing ‘further research.’ This allows them to buy more time to solve the contested issue (for instance, until after elections), although eventually, of course, the findings of the research will come out, possibly at a more convenient moment.

Besides instrumental and symbolic knowledge, there is also a lot of ‘non-utilization’. Much research never finds its way into the policy process, either because it never trickles through or captures attention of any actor (when the enlightenment knowledge creep does not take place, for instance), or because it is consciously ignored. There are many examples of studies that are never utilised because they are seen as too counter-intuitive or framed as ‘fundamental science’, or sometimes because the message they bring is ‘unwelcome’. Policy scientists have defined the strategy of ignoring research as a form of ‘negative feedback’, or conscious efforts to prevent policy change. In some cases, this also involves the discrediting of studies as ‘poor’ or ‘biased’; this involves a form of ‘boundary work’ that we will discuss more in detail later.

However, knowledge utilization is more complex than these different types. It can also involve very different types of knowledge claims that are utilised in the first place. Here it is important to reflect about differences between the various types of dialogue structures discussed earlier. Very different types of knowledge claims can be involved in the different ideal types. The enlightenment model mostly speaks of conceptual and theoretical knowledge, sometimes to be described as ‘fundamental scientific knowledge’, which gradually finds its way into general societal and political discourse through what Weiss (1986) described as ‘knowledge creep’. Think about books and concepts that academics produce, such as ‘assimilation’ or ‘social capital’ or ‘citizenship’, which have eventually become part of migration discourses. The bureaucratic model mostly speaks of information or data provided by scientist, which moreover should fit into existing political and societal discourses in order to be utilised in the first place. Think about the role of ethnic statistics that we discussed earlier. The technocratic model rather speaks of applied knowledge, so knowledge that is directly relevant to policymaking. Finally, the engineering model mostly also speaks of applied knowledge, but then puts primacy on the policymaker to selectively pick-and-choose those strands of expertise that he or she sees fit.

Furthermore, knowledge can be used by very different types of actors. It can be used by policymakers directly involved in policymaking. This is often the case with government-associated advisory bodies or expert committees that have as a primary function to bring together researchers and policymakers. Such contacts often also emerge in a more ad-hoc and informal setting. However, this may also involve other types of actors involved in policymaking in different ways. This includes political actors, such as politicians that like to refer to specific research findings in order to substantiate their policy claims, or political parties that seek to bring in academic expertise more systematically, for instance via party think tanks. Furthermore, interest groups and lobby organisations are actors that will often use academic knowledge, but only if it helps to substantiate their claims.

The timing of knowledge utilization can also matter significantly. Knowledge utilization can take place in very different parts of the policy cycle with different implications. For instance, during the policy stage of agenda setting, knowledge can be used instrumentally to capture attention for specific problems, or symbolically to boost specific policy ideas or actors. Take for instance the role of research in signalling institutional discrimination and subsequently putting this notion on the policy agenda; most European countries now have formal policies regarding institutional discrimination. Similarly, both types of knowledge utilization might apply in the stages of formulating policy proposals and actually taking policy decisions. Knowledge utilization during the stage of implementation often tends to follow a bureaucratic model, where, for instance, the information or data produced by research is used to monitor policy practices and effects. Finally, evaluation is a stage where knowledge utilization often plays a crucial role. However, in that stage as well, utilization can be instrumental to policy change and learning, but also symbolic, for example, the choice for specific evaluation methods that may or may not reflect critically on a policy. For instance, an evaluation of the effectiveness of integration policy can be done in a narrow sense by measuring the effects of a given policy, but also in a broad sense by analysing the perceptions of multiple actors on these policy effects; both designs might deliver very different evaluation outcomes!

Strongly related to the concept of knowledge utilization is the concept of ‘learning’ or ‘policy learning’. Learning is commonly defined as the adjustment of specific beliefs in response to new knowledge, information or experiences. In practice, however, patterns of knowledge utilization can be studied empirically, but whether learning has taken place often remains a more subjective and normative question. For instance, where some claim that a policy may have changed in response to new research, others may claim that external factors such as economic crisis were the cause of policy change and research was only there to substantiate those changes. Hence, the notion of learning or ‘policy learning’ is hardly used in the study of research-policy relations.

16.5 Engaging in Research-Policy Dialogues

The three dimensions of research-policy relations discussed above (dialogue structures, knowledge production and knowledge utilization) may help scholars to make sense of the role they play or can play on the research-policy nexus. A key assertion in this chapter is that there is a strong mutual relationship between these three dimensions. What type of dialogue structure emerges will affect patterns of knowledge utilization as well as knowledge production in the field of research itself. If a dialogue structure puts strong primacy on politics on the research-policy nexus, this will create an opportunity structure for specific knowledge claims and will promote (often more symbolic) forms of knowledge utilization that fit political purposes. Or, when sharp boundaries are established between the role of academics and policymakers, knowledge production may occur more independently and driven by disciplinary and theoretical questions, but possibly resulting in non-utilization in the context of societal and political discourses.

The take home message from this analysis should not be that scholars should refrain from engaging in research-policy dialogues. Yes, as scholars we should always be aware of how such relations affect what we do, the theories and methods we use, the questions we ask, or perhaps even more importantly, the questions that we do not ask. But no, research policy relations do not need to be perverse to our work. The idea of ‘boundaries’ between research and politics meaning that there is a strict separation between both worlds, a sort of impermeable wall, is blatantly simplistic and potentially harmful to the development of a good social scientific understanding of the phenomena we are studying. Especially with strongly policy-relevant topics like migration and integration, relating ourselves to the policy environment can be of vital importance to our (fundamental) scientific research; ignoring the policy context would mean missing out of a very significant factor in how contemporary societies perceive and respond to migration and diversity. Furthermore, there is, even to academics, still value to actually contributing to societal discourses and public responses to migration and diversity.

However, a strong argument has to be made for more reflexivity on the part of scholars when engaging in research-policy dialogues (see also Chap.  6, this volume). Rein and Schon (1994) have made the case for ‘reflective practitioners’ in policy dynamics. This means that actors should always be at the same time aware of the substantive contribution they make as well as the role and position that they take, and the implications that the interaction of roles and substance can have for themselves as well as for others. Scientists should also be ‘reflective practitioners’ when engaging in research-policy dialogues. A better understanding of the interconnections between boundary configurations, knowledge production and knowledge utilization should equip them with the conceptual tools for being ‘reflective’. It is this reflective attitude that enables scholars to be critical not only to policy, but also their own roles, while at the same time not isolating themselves from the social dynamics that are so important to social sciences. It is this reflective attitude that should constitute a ‘social boundary’ between research and politics that does not involve retreat in an Ivory Tower, but rather as a mental or intellectual layer in our social behaviour.

What could be harmful to this reflective attitude is what has been described in the literature as the ‘problem of institutionalization’ (Scholten et al. 2015). When research-policy relations ‘institutionalise’, actors on both sides of the relationship can develop a mutual dependency. For instance, researchers or research institutes may become dependent on funding from the policy environment for their existence, and will thus also be more likely to address those questions that are policy-relevant. The other way around, policymakers will receive knowledge and information from one specific type of actors and probably one knowledge paradigm only. Both factors do not promote a mutual critical attitude, let alone ‘reflexivity’.

In fact, in various countries there have been, over the past, cases where such interdependencies have led to the development of networks or ‘discourse coalitions’ involving specific scholars, institutes and policymakers. For instance, in the UK, a ‘race relations industry’ (Bourne 1980; Small and Solomos 2006) would have developed in response to the close cooperation between research and policy. As a consequence of the close cooperation, the race relations paradigm would have emerged and been reproduced as the dominant integration model in the UK. Similarly, in the Netherlands, a ‘minorities research industry’ would have emerged (Essed and Nimako 2006; Rath 2001) that in a similar way reproduced a specific ‘national model’ of minority integration. According to Rath, this not only reproduced this ethnic minorities model, but also legitimised both academic research to ‘ethnic minorities’ as well as government intervention in the position of these minorities.

Finally, a reflective scholar working on the research-policy nexus should also be aware of social processes within academia that may be triggered by research-policy relations. In the sociology of sciences, various types of so-called ‘boundary work practices’ are defined (Gieryn 1999), which involve patterns of social behaviour that actors develop to define ‘boundaries’ between research and policy as well as prescribe proper ways of mutual interaction. Awareness of such practices will again contribute to the reflexivity of scholars working on the research-policy nexus. One boundary work practice involves the monopolisation of a specific model of doing science or making policies by developing a specific relation with actors and capital in another field (Entzinger and Scholten 2015). For instance, with aid from other fields, actors can strengthen their position within their own field or alter the rules of the game in their favour. Alternatively, boundary work can be aimed at the expulsion of specific actors, by redrawing the boundaries of a field so that specific actors are excluded (for example, depriving researchers of their scientific credibility). Or, boundary work can involve expansion, which occurs when actors that support a specific knowledge paradigm or specific values or ideas about proper research manage to expand that paradigm or those beliefs into other areas as well (ibid, p. 17). Finally, boundary work can be aimed at strengthening the autonomy of research versus other spheres, like politics and policymaking. Autonomy does not mean that fields are not interrelated. Jasanoff (2004) has shown that “keeping politics near but out” forms a very effective strategy for research institutes to strengthen their authority by being involved in policy to some degree.

16.6 Conclusions

This chapter shows how the relationship between research and policy in the field of migration does not only matter in terms of having a societal impact, but can also matter to the development of migration research itself. Perhaps because migration research (in a broad sense involving research to both mobility and migration-related diversity) is a research field rather than an academic discipline, it has been particularly susceptible to developments in the broader social and political setting. The more broadly felt sense of urgency around issues related to migration and diversity, such as the recent refugee crisis in Europe, but earlier also the rise of intra-EU mobility or the increase of labour and family migration in the late twentieth century, has always been an important impetus for migration research. But also the politicization of migration has provided opportunities as well as challenges for migration research.

The central argument in this chapter is that the type of research-policy relationship (or boundary configuration) also matters to knowledge production in the field of migration research as well as to patterns of knowledge utilization in broader society and politics. There is a broad variety in research-policy relations, which has been captured in four ideal types; enlightenment, technocracy, bureaucracy and engineering. Each of these types can impact knowledge production within the field of research in specific ways. For instance, we have seen that under the bureaucratic model a tendency has evolved to collect data or statistics on migration in accordance to very specific national models. Finally, we have seen that the type of research-policy nexus that emerges can also matter to patterns of knowledge utilization. Whereas the technocratic model assumes a rather instrumental form of knowledge utilization, other models can lead to more symbolic forms of knowledge utilization, either to substantiate already existing policy discourses or to legitimise the position of specific actors in the policy process (see also Chap.  13, this volume).

The development of migration research has at various moments been strongly influenced by its relationship to the broader policy environment. We have seen that the politicization of migration throughout Europe has triggered a transformation in types of research policy relations, particularly in the direction of bureaucratic and engineering models that involve more political primacy. We have also seen that, in the past, conceptual and methodological developments in migration research have been at least partly constituted by the perspective of nation states, leading to what is described as ‘national models of integration’. Such national models would, for a long time, have constrained the more comparative and theoretical development of the field. In some countries, even very specific networks or ‘discourse coalitions’ emerged around such national models involving specific policymakers as well as scholars. Finally, we have seen that behind the discourse of ‘evidence based policymaking’, much migration research is used symbolically rather than instrumentally. In some cases, deploying migration scholarship has helped to depoliticise or temporarily remove contentious topics from the policy agenda.

However, observing that research-policy relations has had an important effect on the development of migration research does not per se mean that migration scholars should from now on refrain from engaging in research-policy dialogues. Rather, it requires more reflexivity on the part of scholars in terms of how their relation to the policy setting may also affect their own research. The conceptual toolkit presented in this chapter, derived largely from the sociology of sciences and policy sciences, should help scholars develop higher consciousness on research-policy dialogues. Furthermore, it requires the absence of any form of structural interdependencies between research and policy; the institutionalization of privileged research-policy relations is likely to promote paradigmatic closure and absence of the mutual critical attitude that is required both for policy innovation and good social science research.


  1. Amiraux, V., & Simon, P. (2006). There are no minorities here cultures of scholarship and public debate on immigrants and integration in France. International Journal of Comparative Sociology, 47(3–4), 191–215.CrossRefGoogle Scholar
  2. Bertossi, C. (2011). National models of integration in Europe: A comparative and critical analysis. American Behavioral Scientist, 55(12), 1561–1580.CrossRefGoogle Scholar
  3. Bertossi, C., & Duyvendak, J. W. (2012). Introduction: National models of immigrant integration: the costs for comparative research. Comparative European Politics, 10(3), 237–247.CrossRefGoogle Scholar
  4. Bloemraad, I. (2013). The promise and pitfalls of comparative research design in the study of migration. Migration Studies, 1(1), 27–46.CrossRefGoogle Scholar
  5. Bommes, M., & Morawska, E. (2005). International migration research. Constructions, omissions and the promises of interdisciplinarity. London: Aldershot.Google Scholar
  6. Boswell, C. (2009). The political uses of expert knowledge: Immigration policy and social research. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  7. Boswell, C., & Hunter, A. (2015). Comparing the political functions of independent commissions: The case of UK migrant integration policy. Journal of Comparative Policy Analysis: Research and Practice, 17(1), 10–25.CrossRefGoogle Scholar
  8. Bourne, J. (1980). Cheerleaders and ombudsmen: The sociology of race relations in Britain. Race & Class, 21(4), 431–452.CrossRefGoogle Scholar
  9. Brettell, C. B., & Hollifield, J. F. (1994). Migration theory: Talking across disciplines. London: Routledge.Google Scholar
  10. Caponio, T., Hunter, A., & Verbeek, S. (2015). (De)constructing expertise: Comparing knowledge utilization in the migrant integration “crisis”. Journal of Comparative Policy Analysis: Research and Practice, 17(1), 26–40.CrossRefGoogle Scholar
  11. Entzinger, H., & Scholten, P. (2015). The interplay of knowledge production and policymaking: A comparative analysis of research and policymaking on migrant integration in Germany and the Netherlands. Journal of Comparative Policy Analysis: Research and Practice, 17(1), 60–74.CrossRefGoogle Scholar
  12. Essed, P., & Nimako, K. (2006). Designs and (co) incidents cultures of scholarship and public policy on immigrants/minorities in the Netherlands. International Journal of Comparative Sociology, 47(3–4), 281–312.CrossRefGoogle Scholar
  13. Favell, A. (2003). Integration nations: The nation-state and research on immigrants in Western Europe. Comparative Social Research, 22, 13–42.CrossRefGoogle Scholar
  14. Geddes, A. (2005). Migration research and European integration: The construction and institutionalization of problems of Europe. In Bommes & Morawska (Eds.), International migration research: Constructions, omissions and promises of interdisciplinarity (pp. 265–280). Chicago: Chicago University Press.Google Scholar
  15. Geddes, A., & Scholten, P. (2015). Policy analysis and Europeanization: An analysis of EU migrant integration policymaking. Journal of Comparative Policy Analysis: Research and Practice, 17(1), 41–59.CrossRefGoogle Scholar
  16. Gieryn, T. F. (1999). Cultural boundaries of science: Credibility on the line. Chicago: University of Chicago Press.Google Scholar
  17. Hoppe, R. (2005). Rethinking the science-policy nexus: From knowledge utilization and science technology studies to types of boundary arrangements. Poiesis & Praxis, 3(3), 199–215.CrossRefGoogle Scholar
  18. Jasanoff, S. (Ed.). (2004). States of knowledge: The co-production of science and the social order. London: Routledge.Google Scholar
  19. Jasanoff, S. (2005). The fifth branch: Science advisers as policymakers. Boston: Harvard University Press.Google Scholar
  20. Miller, C. (2001). Hybrid management: Boundary organizations, science policy, and environmental governance in the climate regime. Science, Technology & Human Values, 26(4), 478–500.CrossRefGoogle Scholar
  21. Rath, J. (2001). Research on immigrant ethnic minorities in the Netherlands. In Ratcliffe (Ed.), The politics of social science research (pp. 137–159). London: Palgrave Macmillan UK.CrossRefGoogle Scholar
  22. Rein, M., & Schon, D. A. (1994). Frame reflection: Toward the resolution of intractable policy controversies. New York: Basic Books.Google Scholar
  23. Saharso, S., & Scholten, P. (2013). Comparative migration studies: An introduction. Comparative Migration Studies, 1(1), 1–6.CrossRefGoogle Scholar
  24. Sanderson, I. (2002). Evaluation, policy learning and evidence-based policy making. Public Administration, 80(1), 1–22.CrossRefGoogle Scholar
  25. Scholten, P. (2011). Framing immigrant integration: Dutch research-policy dialogues in comparative perspective (p. 320). Amsterdam: Amsterdam University Press.Google Scholar
  26. Scholten, P., & Verbeek, S. (2015). Politicization and expertise: Changing research–policy dialogues on migrant integration in Europe. Science and Public Policy, 42(2), 188–200.CrossRefGoogle Scholar
  27. Scholten, P., Entzinger, H., & Penninx, R. (2015). Research-policy dialogues on migrant integration in Europe: A conceptual framework and key questions. In Integrating immigrants in Europe (pp. 1–16). Springer.Google Scholar
  28. Simon, P., Piché, V., & Gagnon, A. A. (2015). Social statistics and ethnic diversity. Berlin: Springer.CrossRefGoogle Scholar
  29. Small, S., & Solomos, J. (2006). Race, immigration and politics in Britain: Changing policy agendas and conceptual paradigms 1940s–2000s. International Journal of Comparative Sociology, 8(47), 235–257.CrossRefGoogle Scholar
  30. Thomas, W., & Zaniecki, J. (1926). The polish peasant in Europe and the United States. New York: Alfred A. Knopf.Google Scholar
  31. Thränhardt, D., & Bommes, M. (Eds.). (2010). National paradigms of migration research (Vol. 13). Göttingen: V&R unipress GmbH.Google Scholar
  32. Weiss, C. H. (1986). The circuitry of enlightenment: Diffusion of social science research to policymakers. Knowledge, 8(2), 274–281.CrossRefGoogle Scholar
  33. Wimmer, A., & Glick-Schiller, N. (2002). Methodological nationalism and beyond: Nation–state building, migration and the social sciences. Global Networks, 2(4), 301–334.CrossRefGoogle Scholar
  34. Wittrock, B. (1991). Social knowledge and public policy: Eight models of interaction. In B. Wittrock et al. (Eds.), Social sciences and modern states: National experiences and theoretical crossroads (pp. 333–354). Cambridge: Cambridge University Press.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2018

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Authors and Affiliations

  1. 1.Erasmus University RotterdamRotterdamThe Netherlands

Personalised recommendations