Dissecting the gap : an analysis of the discrepancy between academic research and public policies on the migration-development nexus

(1)

Dissecting the Gap: an Analysis of the Discrepancy between

Academic Research and Public Policies on the

Migration-Development Nexus

Master’s thesis

Marlieke Ruiter

MSc Political Science

Graduate School of Social Sciences

Student ID: 10519327

Supervisor: Darshan Vigneswaran

(2)

Abbreviations

CAMMs Common Agendas on Migration and Mobility DEMIG Determinants of International Migration

EC European Commission

EDF European Development Fund EP European Parliament

EU European Union

EUTF European Union Emergency Trust Fund for Africa

EUR Euro

GAMM Global Approach to Migration and Mobility IOM International Organisation for Migration IOs Immigrant Organizations

MSMEs Micro-, Small- and Medium-size Enterprises NGOs Non-Governmental Organizations

ODA Official Development Assistance

OECD Organisation for Economic Co-operation and Development SINCE Stemming Irregular Migration in Northern and Central Ethiopia SNNPR Southern Nations, Nationalities, and Peoples’ Region

TFEA Treaty on the Functioning of the European Union

Tables and figures

Table 1. Overview and categorization of academic research on research-policy relations p. 19

Table 2. Ideal models of research-policy structures p. 20

Table 3. Overview of policy documents p. 51

(5)

Introduction

Over the last decades, migration and development have increasingly been linked within public policies. At the same time, the relations between both migration and development, including how they interact and influence each other, have been increasingly studied by both researchers and policymakers. In European policy circles, the linkage has been accompanied by the emergence of the idea that a lack of development in migrants’ countries of origin is a ‘root cause’ or ‘driver’ of international migration. The idea seems to have gained ground over the years, evolving into ‘conventional wisdom’ (Castles 2009b; Galbraith 1958). Accordingly, policymakers have1 propagated _{development assistance to address the root causes of migration. The so-called ‘root} cause approach’ is still evident in contemporary migration policies, as it also became visible under the European Agenda on Migration (De Haas, 2007; Clemens and Postel, 2018; Fratzke and Salant, 2018).

However, most migration scholars disagree with this conventional wisdom. These policies seem to be based on a few assumptions that defy academic research findings. Firstly, the assumption that migration and development are negatively related or counterparts. Thereby, a lack of development is perceived as a ‘root cause’ of migration. Secondly, the assumption that through enhancing development emigration from developing countries and regions can be effectively reduced (De Haas, 2007). Accordingly, this notion assumes that the poorest people have to highest tendency to migrate. Opposite this, migration studies demonstrate that the poorest are in most scenarios the least likely to migrate, as they often lack the resources that are needed to make the journey. Furthermore, both people’s capabilities and aspirations to migrate are shown to increase with development (De Haas, 2010; Sen, 1999). Whilst acknowledging that most people migrate to improve their long-term well-being, most migration scholars support a curvilinear relationship between migration and development, which implies that take-off development of poor migrant-sending regions initially coincides with rapidly increasing migration rates (Martin, 1996).

1 _{Stephen Castles used the term ‘conventional wisdom’ as put forward by John Kenneth Galbraith (1958)}

(6)

This illustrates the _{discrepancy between academic research and public policies regarding} the understandings of the relationship between migration and development. At first, this ‘research-policy gap’ might not seem surprising, especially in the policy field of migration, which is highly politicized and securitized. Public policies are not merely evidence-based, as politicians and policymakers weigh different interests and their policy choices are impossibly value-free. At the same time, public policies are expected to rely on evidence to at least a certain extent, as it is desirable that policies meet their publicly stated objectives, regardless of whose interests are represented and which policy outcome is desired. Furthermore, evidence can be used to justify policy choices in public debates and correct public misperceptions (Ruhs, Tamas and Palme, 2019). _{This public task makes the research-policy gap regarding the} migration-development relationship particularly interesting to explore, because public policies that have adopted the ‘root cause approach’ may fail in their stated aims to control and deter migration, as development is likely to increase migration - at least in the short term. Hence, this research-policy gap might result in a policy failure, i.e. that the policy results do not meet their stated objectives.

Moreover, the endurance of this research-policy gap is quite remarkable. The ‘root cause approach’ is not a new phenomenon, as it has been a commonly presented as a ‘smart solution’ to curb migration across Europe since the 1990s and has been a widespread idea in policy circles for the past decades (De Haas, 2007; Clemens and Postel, 2018). Whereas at the same time, this perspective was already challenged by Georges Tapinos in the early 1990s. Since then, Tapinos’ argument has been widely supported by migration scholars, based on the idea of the mobility transition (Castles, 2009; Tapinos, 1990). Although migration will remain an inherently complex and context-dependent phenomenon, migration researchers have become able to identify and understand the dynamics and regularities of migratory processes. The field of migration research seems to have grown significantly over the years, both in volume and quality. Migration theories are increasingly supported by an extensive base of empirical evidence. Whilst on the other side of the coin, European policymakers are becoming more experienced in managing migration. The growing sophistication of migration policies indicates that policymakers’ understanding of migration dynamics has improved over time and enabled them to

(7)

better manage migration to a certain extent (DEMIG, 2018). Altogether, it makes the existence and endurance of the research-policy gap regarding migration and development, constituted by the ‘root cause approach’ in contemporary migration policies, increasingly odd and relevant to research.

Although the research-policy gap has not gone unnoticed to academics, it has not been thoroughly studied to date. A large group of studies has aimed to analyse and conceptualise the gaps between the two spheres of academia and policymaking in the field of migration, yet the particular research-policy gap regarding migration and development has not been dissected and explained. A few migration scholars have also diagnosed the ‘root cause approach’, addressing that it has become ‘conventional wisdom’ in policy circles (Castles, 2009a; 2009b; De Haas, 2007). More recent academic endeavours have been critical of the ‘root cause approach’ and how it is adopted under the European Union’s Agenda on Migration since 2015 (see e.g. Clemens and Postel, 2018; Fratzke and Salant, 2018; Kipp, 2018). These studies have all addressed the discrepancy, but did not include a thorough analysis of the research-policy gap or the understandings of policymakers. Instead, policymakers’ understandings of migration and development seems to be either neglected or taken for granted. Therefore, this thesis aims to thoroughly analyse and dissect the research-policy gap. It focuses on the ‘root cause approach’ within the external dimension of European Agenda on Migration, with the Trust Fund for Africa as its most significant tool. Through a manifest and latent qualitative content analysis of both academic research and public policy documents, I will analyse the research-policy gap on the relationship between migration and development, in order to detect possible explanations for its existence. It aims to answer the research question: ‘What explains the existence of the research-policy gap on migration and development?’

This thesis is organized as follows. Chapter 1 reviews the academic literature on the migration-development debate and the research-policy gap in the migration-development field. In Chapter 2, I will highlight useful conceptualisations of research-policy relations from existing research and use this to form my research framework. It formulates three possible explanations for the existence of a research-policy gap, followed by three sub-questions to guide this thesis. Subsequently, the research design and methodology are discussed. Chapter 3 and 4 include the

(8)

main analysis. In Chapter 3, I will identify and review the most relevant academic research findings regarding the migration-development relationship. In Chapter 4, I present the results of the content analysis of policy documents, in order to dissect the ‘root cause approach’ and the existence and form of the research-policy gap. I will finish with my final conclusions and remarks, including some recommendations for further research.

(9)

Chapter 1 Literature review

1.1 Diagnosing the research-policy gap

This section reviews the literature on the gap between public policies and academic research concerning the developmental drivers of migration and the relationship between migration and development more broadly. The research-policy gap, as addressed in this thesis, has not been thoroughly studied thus far, but has been diagnosed by some migration scholars. Furthermore, migration research has reflected substantially on earlier migration and development policies and the role of states in governing migration more generally. Since states and their (migration) policies can significantly influence migrants’ lives, the role of the state has logically been addressed by migration scholars. The first section reviews the academic reflections on restrictive immigration policies and the second section reviews what has been done with regard to analysing alternative ‘smart solutions’ to deter migration.

Academic reflections on the ‘quest for control’

Since the mid-1970s, immigration from developing countries, in particular immigration of low-skilled workers and asylum seekers from poorer countries, has been increasingly perceived as a problem in need of control in most Western liberal democracies (De Haas, 2007; Castles, 2000). As a result, the opposition to this immigration “burden” has risen, which resulted in aims to restrict and deter immigration and stimulate return migration. Since the 1990s, the academic debate turned to the “puzzling paradox” of how large-scale settlement has taken place, even though Western liberal governments and the public considered such immigration undesirable (Bonjour, 2011). As most migrants arrived through legal channels and were given several rights over time, the question soon became “why liberal states have _{accepted unwanted migration”} (Joppke, 1998, emphasis added). Academic reflections on states’ immigration policies have been fairly critical, which exemplifies early discrepancies between the understandings of academics

(10)

and policymakers on the drivers and governance of migration. Immigration policies were perceived as either failing to meet their objectives or liberal states were unable to restrict their policies to their desire (Castles, 2004; Czaika and De Haas, 2017; Bonjour, 2011). According to some scholars, liberal democratic governments have lost their control over migration to domestic courts, while others argue that global human rights regimes limit governments’ abilities to control migration. Another group of scholars have argued that the “control gap” is constituted of other constraints for policymakers, such as differing interests that need to be taken into account (Messina and Lahav, 2006; Freeman, 1994)._{Most of the debate’s questions remain relevant. For} example, if there are limits to states’ regulatory capacities in this field, and if so to what extent and of what nature (Bonjour, 2011). However, since the immigration flows to Europe since 2 1945 were largely unprecedented, early migration policies (i.a. guestworkers programmes, entry policies and family reunification) were designed in the absence of an extensive research base. According to Saskia Bonjour, this left room for decision-makers to opt for assumptions that fit their policy preferences best, since “assumptions are born in the interplay between incomplete information and political preference” (idem, p. 113). As a result, a lot of these policies were “flawed” and facilitated unforeseen large-scale settlement, because states failed to anticipate the development and dynamics of migration processes.

Academic reflections on ‘smart solutions’ to control migration

Two renowned migration scholars, Hein de Haas and Stephen Castles, have explored the changing views on migration and development and analysed alternative ‘smart solutions’ to deter migration. According to De Haas (2006a; 2007) the “failure of restrictionism” confronted states with their inability to control migration through restricting legal channels, insofar as they were able to do so. In response, De Haas states, policymakers, politicians, development NGOs - and academics - have advocated ‘smart solutions’ to deter future immigration, namely with trade liberalisation and development aid policies in the same vein as the root cause approach, or as De Haas phrases it “development instead of migration policies” or “aid and trade instead of

2 _{See Bonjour (2011) for an overview of this debate, which is commonly known as “the control gap} debate.”

(11)

migration policies” (De Haas, 2006a; 2007). In his paper, De Haas aims “to argue why such3 policies are bound to fail, primarily because they seem to be based on _{erroneous assumptions} about the highly complex ways and often positive ways in which migration and development processes are reciprocally related. Although this insight is well-established in academic discourse, erroneous notions on development-migration linkages still persist in many policy circles” (De Haas, 2006a, p. 5, emphasis added). Thus, in this quote, De Haas acknowledges the existence of a gap between the insights of academics versus those of policymakers - which was brought to his attention by Stephen Castles. Furthermore, De Haas stresses that the commonly presented ‘smart solutions’ to reduce migration pressures (i.e. aid, trade and temporary or circular migration programmes) are based on “an inaccurate analysis of the dynamics that drive migration” (idem, p. 5). According to De Haas, there is a gap on the development-migration linkages between the well-established insights in academic discourse and erroneous assumptions in policy circles. Although he does state, in the beginning of his paper, that academics and development NGOs have also advocated smart solutions to deter migration, which suggests that there are (or were) researchers that believe(d) the erroneous notions underlying the root cause approach. Regarding the understandings of policymakers, De Haas talks about “beliefs” of policymakers and their “inaccurate analysis” and “erroneous assumptions” underlying the development instead of migration policies. Hence, De Haas seems to suppose that policymakers do not understand the dynamics that drive migration and sincerely belief they can deter migration through development. Accordingly, De Haas seems to assume that these policies are sincere in their aims to bring about development, even though it is meant to be “stay-at-home development” (Böhning, 1994) in order to deter migration. As he concludes that “advocates of stay-at-home development policies seem to be right for the wrong reasons” (De Haas, 2006a, p. 31).

In 2009, Stephen Castles stressed that it has become conventional wisdom in both popular and political discourse to argue that promoting economic development has the potential to reduce migration (Castles, 2009a; 2009b). In his articles, Castles elaborates on a meeting with

3_{See for example Schiff (1996) and Martin (2002) for reviews on trade liberalisation policies and Stalker} (2000) and Böhning and Schloeter-Paredes (1994) for reviews on development aid policies.

(12)

a former senior European Union functionary in 2008, in which it became evident that the ex-official, who had been involved in the EU’s external affairs, had thus far strongly believed that development policies can drastically reduce international migration insofar as they are able to tackle the root causes of migration. According to Castles, this anecdote tells us that this “strongly held belief” is broadly shared among politicians, officials and the public (Castles, 2009b, p. 442). It does not become clear whether Castles has met with other policymakers. Castles continues by stating that the argument “that development policies cannot quickly reduce international migration, because a higher level of development brings more mobility, not less” was raised over 20 years ago by Georges Tapinos (1990) and “is now one of the few things that virtually all migration scholars agree with” (Castles, 2009b, p. 442). Thus, although migration scholars often do not agree, Castles states that there is considerable consensus on this idea, which has been circulating in academic circles for over 20 years. After stating this, Castles does not return to the question why politicians and officials “still strongly believe” in the root cause approach, instead he reflects on the research carried out by migration scholars:

“...our research findings are apparently not being listened to. If decision-makers still do not understand one of the most basic facts about the migration/development relationship, then either there is something wrong with the research carried out by migration scholars, or we have failed to bridge the gap with decision-makers – or both. In fact I believe that it is both.” (Castles, 2009b, p. 442).

This excerpt has several interesting elements. Firstly, the assumption that decision-makers do not listen to research findings and still do not understand the argument (here even referred to as a “basic fact”) about the migration-development relationship. Secondly, the argument that migration scholars are for a large part responsible for this gap, either through the content, quality or dissemination of their research. According to Castles, the gap is caused by the fragmentation and narrowness of migration studies that has resulted in a lack of shared concepts and questions, and the absence of an accumulation of knowledge. Furthermore, he states that migration scholars have not been very good at getting this important issue over to non-specialists, i.e. policymakers

(13)

(idem). In sum, Castles diagnosed the gap between research and policy and argued, in the same vein as De Haas, why these policies will not quickly reduce migration. Castles’ inquiry (2009a; 2009b) into the limitations of migration studies has led him to conclude that this gap is a result of the fragmentation and narrowness of the existing base of studies and the failure of researchers to bridge the gap with policymakers. Regarding the beliefs of policymakers, Castles seems to be convinced that most politicians, policymakers and the public are either not aware or do not understand the argument about the migration-development relationship. As stated earlier, it is not clear whether Castles has made any further inquiries into the understandings and beliefs of policymakers. However, at the end of his article, Castles adds more nuance to his position. First, he states that there is considerable evidence that improvements in living standards and reductions in violence actually create the conditions for people to migrate (UNDP, 2009) and argues that the consequences of development instead of migration policies were spelled out convincingly the 2005 Report of the Global Commission on International Migration (Castles, 2009b). According to Castles, these policy reports indicate that some government officials are in fact aware of the complexity of the migration-development relationship, although key decision-makers might not. It also illustrates that researchers’ messages are not at all times ignored, which should reinforce researchers’ determination to further develop their analysis and overcome the barriers with other researchers and policymakers (idem).

Academic reflections on the migration-development debate

Both De Haas (2010a; 2010b; 2012) and Castles (2009a; 2009b) have also reflected on the migration-development debate since the 1950s and the interplay between research and policy within this debate. The debate has shifted remarkably over the past half century, from predominantly optimistic to pessimistic views and back again to optimistic views, moving back and forth like a “pendulum” (De Haas, 2012). In his critical review of the debate in both academic and policy discourses, De Haas identified the shifts from optimistic to pessimistic views, particularly regarding the effects of migration on development in sending countries. According to De Haas, the “pendulum” is part of more general shifts in development theory, and hence rooted in deeper ideological and paradigmatic shifts (idem). His inquiry is mostly a

(14)

reaction to the “neo-optimism” about migrants’ role in development that regained ground in the 1990s and early 2000s, which De Haas perceives as naive, overly optimistic and based on neoliberal views instead of the upsurge in remittances (idem, p. 10). Thus, De Haas has demonstrated how academic discourses have been influenced by policy discourses and broader ideologies.

According to Castles, the migration-development debate appears to have consisted of two separate, but intersecting, discourses: an academic and a policy debate. Castles states that it is important to look at the interplay between economic and political interests and the development of migration research, because remarkable turnarounds in the academic debate might reflect states’ interests_{(Castles and Van Hear 2010; Castles 2009b). Reflecting on the interplay between} the debates, Castles argues that migration research, and social science research more generally, has been readily available to support public policy initiatives. This does not imply that scientists have been “unduly influenced by those in power” but does point out that scientists can be steered in the direction of political needs (Castles, 2009b, p. 453). Castles also illustrates that policymakers can pick-and-choose from research findings, because of the controversies and inability to agree within migration research. Whether the recent turnaround to optimistic views of the migration-development relationship reflects a “scientific revolution” _{(referring to Kuhn,} 1996) or a shift in political needs, remains a question that Castles finds difficult to answer. However, Castles concludes, “it is really up to social scientists working on migration and development” to develop their approaches and not yield to short-term political needs (Castles, 2009b, p. 453).

(15)

1.2 Analysing the ‘root cause approach’ in the European Agenda on

Migration

Several recent academic endeavours have reflected on the root cause approach and how it is adopted and implemented under the European Union’s Agenda on Migration. This section summarizes their main findings and reflects on their contributions. Susan Fratzke and Brian Salant (2018) stress that the twin policy areas of migration and development have become inexorably connected in the minds of political leaders worldwide, especially in countries that have recently experienced increases in unmanaged migration, which raises their concerns. In 4 their policy brief, they reflect on the research and experience in the migration and development field, arguing that the relationship between migration and development is anything but simple.

Fratzke and Salant do not seem to have thoroughly analysed the understandings of policymakers. They advise policymakers to reconsider their ‘root cause wisdom’ in order to achieve the objective of restricting migration, because the relationship is much more nuanced than acknowledged by policy discourses (Fratzke and Salant, 2018). Furthermore, they acknowledge that political leaders and policymakers are likely to be focused on the next election cycle and short-term migration management priorities, and may find “development assistance to be a blunt tool for reshaping migration patterns” (idem, p. 2). In the end, Fratzke and Salant argue, this short-term focus might negatively influence development goals, and they stress that the strength with which migration and development have become linked raises concerns (idem). Fratzke and Salant stress that evidence suggests that the migration-development relationship is nuanced and extremely context-specific. In another paper, they have assessed research since 2002 on how economic conditions, livelihood opportunities and development interventions affect migration. Fratzke and Salant conclude that little specific research has been5 done on the effects of development policies on migration patterns, plus very few development

4 _{Policy brief of Migration Policy Institute, January 2018. The MPI is an independent, nonpartisan,} nonprofit think tank. The publication results from a partnership between MPI and the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, supported by the German Federal Ministry for Economic Cooperation and Development (BMZ).

5 _{see Fratzke and Salant ‘Understanding the impact of livelihood opportunities and interventions on} migration patterns’ (2018) UK Department for International Development_.

(16)

interventions monitor migration outcomes (idem). This might indicate that academic research has not been policy relevant to date. In sum, Fratzke and Salant basically provide a summary of past research and policy experiences, including some policy implications; as it is a policy brief. Fratzke and Salant have done an excessive research assessment, so their views on the existing literature are not unfounded. However, apart from briefly mentioning some possibilities regarding policymakers’ intentions, they do not make any further inquiries into analysing what has caused this discrepancy between research and policy wisdom.

David Kipp (2018) reviews the European Agenda on Migration and the EU Trust Fund (EUTF) for Africa, in order to come up with recommendations for the German government, as the largest contributor to the Fund among EU member states. Kipp perceives the Agenda as a 6 direct response to the increased refugee and migratory movements into the EU since 2015 and the failure of internal cooperation between the member states. This implies that the aim of deterring migration is higher on the policy agenda than enhancing development. According to Kipp, the influx revealed deficits in the common migration and asylum policy, yet little has been achieved with regard to sharing responsibility. Due to the differing interests of member states, agreement to reform the Dublin system seems impossible. However, as most member states do agree to reduce arrivals, the external dimension brought opportunities (Kipp, 2018).

The research paper continues to review the Trust Fund, as it is the new important instrument under the Agenda. It analyses its goals, funding, governance structures and practice, upon which it concludes that despite its broad objectives, the main purpose of the EUTF seems to be to secure the cooperation of third countries in reducing irregular migration and refugee flows and to take back irregular migrants. Kipp illustrates this by pointing to a progress report which “attributes declining arrivals in Europe partly to EUTF measures in the area of migration management and security” but does “little in the way of evaluation of development initiatives” (Kipp, 2018, p. 11). Thus, Kipp makes more rigorous statements regarding the aims and objectives of European policymakers. This seems to be based on a thorough analysis of policy documents and policy practices under the EUTF. Although this is a valuable contribution to the

6 _{Research Paper of}_{Stiftung Wissenschaft und Politik (SWP) German Institute for International and} Security Affairs, December 2018. It was prepared within the framework of a research project funded by the German Federal Ministry for Economic Cooperation and Development (BMZ).

(17)

analysis of the objectives of the root cause approach, it does not include an analysis of research findings or the understandings of policymakers regarding the migration-development relationship.

Michael Clemens and Hannah Postel (2018) have noted that development agencies in Europe acquired a renewed mandate in response to the influx of irregular migrants in 2015, namely to deter migration from poor countries. They highlight that these efforts did not come out of nowhere, but follow decades of recommendations from other policymakers in Europe (Khoudour-Castéras, 2009; Clemens and Postel, 2018). Their article reviews existing evidence on whether development aid has historically deterred emigration from poor countries, focusing on two requirements for the success of these policies: the effects of aid on development and the effects of development on migration. Hence, they consider the relation between aid and migration, and the effects of development assistance more specifically. Regarding the existing base of research, Clemens and Postel state that there are few cross-country studies testing the overall relationship between aid and migration, and that those few fail to offer clear evidence that aid has substantially deterred migration on average. According to Clemens and Postel, these studies face important challenges in “interpreting the coefficient estimates on aid as the policy impact of an increase in aid, due to inherent limitations on identifying causal relationships in cross-country data” (Clemens and Postel, 2018, p. 685). Clemens and Postel take a granular approach to assessing aid targeted at ameliorating the root causes of migration, which has not been done thus far, and compiled an index of migration-relevant aid, drawing on specific programme types detailed for the EU Trust Fund. They conclude that _{the evidence that aid can} sustainably deter migration from developing countries is weak at best. Regarding policymakers’ role, Clemens and Postel argue that aid tends to follow donors’ geopolitical concerns (Kuziemko and Werker, 2006; Faye and Niehaus, 2012). In their lessons for aid policymakers and researchers, they recommend to look beyond efforts to deter migration and increase efforts to shape migration to be beneficial for all. According to Clemens and Postel, “current aid efforts around the world have devoted essentially _{none of their portfolio to supporting innovative ways} to shape rather than deter migration.” (Clemens and Postel, 2018, p. 688, emphasis added). _To exemplify this, they stress that none of the projects under the EU Trust Fund have the goal of

(18)

creating mutually beneficial labor mobility channels between Africa and the rest of the world. Furthermore, they stress the need for improved documentation of aid programming to help researchers and policymakers to understand this phenomenon, because the current public documentation is too high-level (idem).

1.3 Knowledge gaps

The academic contributions of De Haas and Castles have been incredibly important for the diagnosis of the ‘root cause approach’ and the gap between research and policy on the migration-development relationship. However, they seem to have been occupied by countering the conventional wisdom in policy circles, rather than to analyse were the policy wisdom comes from and which considerations are truly underlying the root cause approach. Furthermore, De Haas and Castles have contributed significantly to the mapping of the migration-development debate and the interplay between research and policy within this debate, yet these inquiries were focused on high-level paradigmatic shifts within the discourses. An in-depth, thorough analysis of the research-policy gap, and in particular the insights of policymakers, is still missing. Castles does reflect on the existing base of migration research to account for the gap, but seems to leave the understandings of policymakers untouched. Regarding the role of policymakers, Castles and De Haas both seem to assume that most politicians, policymakers and the public are either not aware of, or do not understand the argument about the migration-development relationship.

The recent academic contributions have reflected on earlier evidence and approaches and analysed the root cause approach under the European Union’s Agenda on Migration. All highlighted and reflected on some peculiar elements of the policy implementation and practice under the Trust Fund, which provide meaningful insights into the understandings and interests of policymakers. However, although these contributions include extensive research assessments, the analysis of policymakers’ insights is limited, as these contributions do not primarily focus on analysing and explaining the research-policy gap.

(19)

Chapter 2 Research framework

2.1 Conceptualising research-policy relations

This section provides an overview of the existing literature and conceptualisations of research-policy relations. At the same time, it constitutes the research framework for this thesis project, as I will use the literature to conceptualise the research-policy gap. Whilst most research on research-policy relations presented here is conducted within the field of migration, some studies have focused on research-policy relations more generally. The existing literature has focused primarily on research-policy dialogues in the field of migration and integration. Hence, the research-policy gap regarding migration and development has not been covered by this strand of literature, creating a knowledge gap in the literature. However, the contributions covering migration and integration research remain useful for analysing and conceptualising the research-policy relations regarding migration and development, as it takes largely place within the same policy field of migration. Although this research project is modest in its aims, as it is neither able to sufficiently nor completely grasp the research-policy relations concerning migration and development, the concepts and theories provided by the earlier academic work on research-policy relations provide useful insights that I can adopt as tools for this project. Although my analytical focus is not on the policymaking process and research-policy dialogues directly, but rather on the policy and research output, the phenomenon under study is the research-policy gap. Hence, these theoretical insights are employed for my research framework.

The role of knowledge in public policymaking

The role of knowledge and expertise in public policymaking has been frequently studied within academic research. From the mid-1970s onwards, there has been given particular attention to how knowledge is utilized for policymaking purposes. Today, there exists an extensive research 7

(20)

base on the interactions between research and policymaking and how this relationship has evolved over time. The interest is not solely from academic spheres, as both the public as well as policymakers and politicians have increasingly demonstrated their interest in the role of knowledge in policymaking, building a case for “evidence-based policymaking” (Entzinger and Scholten, 2019). In order to structure the existing literature and conceptualise research-policy dialogues, Han Entzinger and Peter Scholten (2019) have come up with categorization of three strands of literature, which I will use to organize this section and which I will refer to throughout this thesis. The categorization is based on the research focus, which is either (i) research-policy dialogue structures; (ii) knowledge utilization or (iii) knowledge production. Table 1 provides an overview of this categorization, including the existing body of research within each category.

Table 1. Overview and categorization of academic research on research-policy relations

Category Research focus Existing literature

Research-policy dialogue structures

Formal and informal arrangements through which knowledge, decisions on knowledge production and

relevance of knowledge for policy are exchanged

Bommes and Morawska (2005); Florence and Martiniello (2005); Geddes (2005); Penninx (2005); Thränhardt and Bommes (2010); Scholten (2011)

Knowledge utilization Cultures and practices of knowledge utilization in policymaking processes, analysing the role that has been assigned to researchers and the function attributed to knowledge and research

Boswell (2008; 2009; 2019)

Knowledge production Cultures of knowledge production and the impact of policy on research

Favell (2003); Penninx (2005); Vasta and Vuddamalay (2006); Thränhard and Bommes (2010) Source: own elaboration based on Entzinger and Scholten’s categorization (2019).

(21)

Research-policy dialogue structures

This category focuses on the formal and informal arrangements through which knowledge, decisions on knowledge production and relevance of knowledge for policy are exchanged. Research-policy dialogues are important for the relation between research and policymaking. Several factors may determine what kind of knowledge is transferred and how. For example, the organization of the exchange, the location, the involved actors, the types of knowledge and communication and the discussed issues. A number of ideal models of research-policy structures have been defined and brought forward by different scholars (Hoppe, 2005; Scholten, 2011). These are four models, that can be categorized based on two features: (i) whether a primary role in policymaking is given to research or politics and (ii) how strict the boundaries are between the role of the researcher and the role of the policymaker. Table 2 below provides an overview.

Table 2. Ideal models of research-policy structures

Primary role for research Primary role for politics

Strict boundary between researchers and policymakers

Enlightenment model

- ‘Speaking truth to power’

- Strict boundaries between research and policy

- Indirectly determining how policymakers interpret and act upon policy problems

Bureaucratic model

- Research provides data and ‘facts’ to policymakers

- Policymakers develop policies and reach decisions

- Sharp fact–value dichotomy Loose or less strict

boundary between researchers and policymakers

Technocratic model

- ‘Experts’ are more directly involved in policymaking, they do more than just provide knowledge

- Experts also frame policy problems and develop solutions

- Experts as policymakers

Engineering model

- Researchers have more far-reaching role in policymaking

- Politics keeps its primacy and is at liberty to select (‘pick-and-choose’) those strands of expertise that it sees t.

(22)

These four models are primarily distinguished by the function given to research and knowledge in the process of policymaking. However, these functions also translate into the forms of the dialogues. Hence, these models can also serve to map differences between forms of dialogues. Differences can exist for example in structures or in degrees of institutionalization. A stronger institutionalization of research-policy relations can imply a substantial influence of research on policymaking, that would fit within the logic of the technocratic model. Another example can be a research-policy dialogues with more informal and personal networks, which can result in a strong primacy for politics (i.e. the bureaucratic model).

Knowledge utilization

The category of knowledge utilization focuses on the cultures and practices of knowledge utilization in policymaking processes, analysing the role that has been assigned to researchers and the function attributed to knowledge. Institutions and dialogue structures form the exchange of knowledge between researchers and policymakers, and facilitate the transfer of knowledge. After knowledge is transferred, i.e. policymakers have access to information, knowledge can be selected and used in different ways. Broadly, there are two types of knowledge utilization: (i) instrumental utilization, which means that expert knowledge is directly taken as input for policymaking and instrumentally used and (ii) symbolic utilization, which means that research is not directly taken as input for policymaking but rather has a symbolic function in supporting policies (Boswell, 2009; 2019; Entzinger and Scholten, 2019). Christina Boswell (2008; 2009; 2019) has distinguished two different types of symbolic utilization: substantiating and legitimizing. In the first case, knowledge can be used to provide authority to policies (which have already been decided) by _{substantiating} _{these through selected relevant and supportive} expert knowledge. In the second case, expert knowledge and research is used for a more broader

legitimization of policies and policy institutions. Knowledge does not need to substantiate the adopted policy approach per se, as this form concerns the mere symbolic act of having knowledge to claim authority over a particular policy domain or issue. Thus, both are symbolic uses of knowledge to claim authority, one focusing on selecting evidence to support policies

(23)

rather than the policymaking process, and the other focusing on building a reputation in general (idem).

Knowledge production

The category of knowledge production analyses both the quality and production of expert knowledge. It assesses how policy interests and research-policy dialogues have affected knowledge (production). Thus, focusing on the other side of the research-policy spectrum, researchers have looked at knowledge production and its interactions with policy dialogues, assessing in what ways and to what extent knowledge (production) has been affected.

Research-policy dialogues can facilitate certain opportunity structures in the research field for some researchers, programmes and institutions to put forward their work and influence policymaking (Penninx, 1988; Entzinger and Scholten, 2014; Jasanoff 2005). Furthermore, institutionalized research-policy dialogues can affect the structural characteristics of migration research as a research field. For example, the extent of consensus or fragmentation (Entzinger and Scholten, 2019). Various scholars have stressed that research-policy dialogues can substantially impact methodological, theoretical and disciplinary developments within knowledge production. Particularly, the argument that research-policy dialogues have favoured the nation-state as a ‘constitutive frame’ is put forward. According to Thränhardt and Bommes (2010) this impedes the theoretical development of migration and the rise of a more critical approach to the role of the nation-state. For example, the issue of integration has been addressed as a problem to be solved from a nation-state perspective, rather than from a scientific perspective (Favell, 2003; Thränhardt and Bommes, 2010). Wimmer and Glick Schiller (2002) have also referred to comparable biases and pointed at the ‘methodological nationalism’ in migration research. These scholars argue that migration, immigration and integration should be conceptualised and theorised from a more neutral scientific perspective. Altogether, Figure 1 provides a visualization of the three main aspects of research-policy relations. As it is borrowed from Entzinger and Scholten’s conceptualisation, it is focused on the field of immigrant integration. However, this conceptualisation can also be applied to the field of migration and

(24)

development. This thesis analysis for the most part situated on the outsides of both circles, since it analyses research and policy output, and to what extent the gap exists between these.

Figure 1. Overview of the three main aspects of research-policy relations

Source: Entzinger and Scholten (2019).

Possible explanations for the research-policy gap

Building on the existing knowledge and experience, it is possible to survey some explanations for the existence of research-policy gaps in general, and the gap regarding the developmental drivers of migration more specifically. Broadly defined, there seem to be three possible explanations for the existence of a research-policy gap. Following Boswell’s (2008; 2009; 2019) and Entzinger and Scholten’s (2019) conceptualisations, these explanations of the research-policy gap can be divided over the domains of research-policy dialogue structures, knowledge production and knowledge utilization. Hypothetically, if a research-policy gap does not exist, there is sufficient expert knowledge and evidence, which is properly transferred and instrumentally used by policymakers. Thus, the processes of knowledge production, knowledge transfer and knowledge utilization all function sufficiently in order to achieve evidence-based policies. However, in the case of a research-policy gap, a deficiency or ‘error’ can be found in each part of the process. The three broad explanations account for possible deficiencies in each

(25)

domain, but are not mutually exclusive, since deficiencies may be found in more than one domain and affect each other through the interplay between knowledge production, research-policy dialogues and knowledge utilization.

Explanation 1: Knowledge is produced and transferred, but not used instrumentally

The first possible explanation concerns the domain of knowledge utilization. In this scenario, there is sufficient expert knowledge, which is sufficiently disseminated and transferred, but policymakers either consciously or unconsciously not use that knowledge instrumentally. Policymakers might unknowingly ignore evidence because they strongly hold another belief or knowingly ignore evidence because of political considerations. In the policy area of migration, it is highly possible that there exists a substantial gap between expert knowledge and public opinion, through the high levels of politicization and securitization of migration in media coverage and the public debate. This makes certain policy decisions infeasible, as politicians’ and policymakers’ room for manoeuvre is limited, even if the research is commissioned by the ministry itself. Hence, electoral pressure and political considerations might prevail over expert knowledge. In this case, knowledge might still be used in a symbolic function, either through legitimizing the governing institutions and their policies and claim authority by the mere symbolic act of having knowledge or through selectively presenting substantiating expert knowledge to support the policies.

Explanation 2: Knowledge is not transferred

The second explanation can be a lack of what Boswell phrases as “organizational capacity.” According to Boswell, it is possible that policymakers are unable to make effective use of expert knowledge, because they lack organizational capacity (Boswell, 2008). In this scenario, there is sufficient expert knowledge, but policymakers are not able to use that knowledge instrumentally. However, this scenario can also be applied more broadly to research-policy relations, approaching _{it as a deficiency within the transfer of knowledge. This means that a lack of}

(26)

organizational capacity can also be found within the spheres of academic researchers or research-policy dialogue structures. Although there might be sufficient expert knowledge, it is not sufficiently transferred or used because of a lack of organizational capacity of policymakers, researchers or other actors involved in research-policy dialogues. This does not concern _the relevance or content of academic research or the willingness of policymakers to use knowledge instrumentally.

Explanation 3: Lack of knowledge

The third explanation can be found within the sphere of knowledge production. Following Boswell (2008), this explanation is a lack of relevant knowledge. This shifts the responsibility to the knowledge producers, as the research produced may be too abstract for resolving policy issues or not presented in an accessible manner. There might also be other limitations to the existing base of academic research. This explanation concerns the content of academic research findings, and not necessarily the organization or communication. However, this explanation is linked with the second explanation, as the production of academic research can also be influenced by dialogue structures. These explanations translate into three sub-questions, that will help to answer the research question of this thesis: What explains the existence of the research-policy gap on migration and development?

Sub-question 1: Is there sufficient expert knowledge?

Sub-question 2: Is expert knowledge sufficiently transferred? Sub-question 3: Is expert knowledge used instrumentally?

In order to provide answers to these questions, I will use an in-depth analysis of academic research and public policies. I will first analyse the limitations and quality of the existing base of academic research covering the relationship between migration and development, and the developmental drivers of migration in particular. This analysis will focus on both the content as well as the dissemination of expert knowledge. Secondly, I will analyse migration policies in

(27)

detail, assessing to what extent knowledge is transferred and used. The analysis will focus mostly on policies on paper, i.e. policy documents.However, it is unfortunately not possible to completely answer my research questions, as it is beyond the scope of this thesis to explore the research-policy relations in all its aspects. An analysis of the daily practice of research-policy dialogues on different governance levels is not included. However, the in-depth analysis of both research and policies will enable us to retrieve some insights of research-policy dialogues. This thesis should therefore be regarded as a first endeavour into a more thorough analysis of this remarkable research-policy gap. In the following section I will elaborate on the methodological choices I have made in order to conduct the empirical exploration of the thesis.

2.2 Research design

Scope and concepts

(i) Academic research

The relationship between migration and development is studied by scholars from different academic disciplines and backgrounds. The main focus of this thesis is on academic research conducted by ‘migration scholars’. Migration scholars are conceptualised as scholars with a specific research focus on migration, as migration studies is an interdisciplinary field of study. Migration studies draws on different disciplines, such as anthropology, sociology, economics, political science and history. In general, the effects of development on migration have been particularly studied by migration scholars, whereas the broader relationship between migration and development, including the effects of migration on development, is mostly covered by development studies and ‘development scholars.’ Therefore, this thesis’ analysis focuses8 primarily on, but is not confined to, academic research findings from migration scholars. As it is nearly impossible to cover all research conducted, both within and outside of the academic sphere, the analysis covers the most relevant academic works. Among the academic research studied, there is not necessarily a geographical focus. Migration scholars both originate from and

8 _{Development scholars are conceptualised in the same vein as migration scholars. Development and} migration studies are both interdisciplinary fields of study, brought together by the same object of study. Hence, development scholars are mostly occupied with studying development.

(28)

cover almost all continents. As for the considered time period, the analysis roughly covers the period from the 1970s until the early 2010s. This time period is not strictly set, as it is based on the emergence of the field and the most relevant publications prior to the adoption of the European Agenda on Migration.

The term ‘research’ is used as shorthand for concepts and theories, data, and empirical ndings, as well as ideas and perspectives. Thus, the approach to conceptualising and studying research is fairly broad and covers a range of different types of knowledge. Throughout this thesis, the terms ‘research’, ‘researchers’, ‘scholars’, ‘academics’, ‘academic’ or ‘scholarly research’ i.a. are used interchangeably and refer to research conducted within the academic field or research projects outside of the field that are conducted by academic researchers. The terms ‘researchers’ and ‘research’ are used to a lesser extent, as this may be confusing. Of course, there is a lot of research conducted outside of the academic sphere, but this is mostly not covered in the analysis. These lines and distinctions might be fluid and unclear in the field, since there is a lot of research done ‘in between’, conducted by ‘boundary organizations’.

(ii) Public policies

The public policies under consideration are policies of migrant-receiving countries and supranational entities as the European Union that explicitly or implicitly link migration and development, and thereby aim to tackle development in order to reduce migration. These are first and foremost migration policies, since they have the aim to reduce migration pressure and development policies focus primarily on development-related goals. However, the policies might also fall under the mandate of development departments. As migration and development goals are linked, there naturally might be some overlap between the mandates of both policy departments. The focus is on contemporary policies, those that are currently in place, or adopted over the last years.

The geographical focus of this thesis is on Europe, as the policies under analysis within the research-policy gap fall under the European Agenda on Migration. This is merely the case for the ‘policy side’ of the research-policy gap, as earlier mentioned, the analysed academic research is not confined to European spheres or African-European migration. This focus can be explained

(29)

through several reasons. First of all, the ‘root cause approach’, and hence the research-policy gap, seem most evident in the public policies of migrant-receiving states. Secondly, Europe makes an interesting, important and unique case for studying the relation between migration and development, and how receiving states aim to govern this. Europe is a relatively high-income, migrant-receiving continent, surrounded by less-developed and unstable regions and continents. Europe’s geographical location facilitates multiple opportunities for entry through ‘irregular’ pathways, in particular for migration from Africa and the Middle East. Several possible routes over land and sea exist, whereas other high-income continents seem less easier to access. Furthermore, recent developments have made the European case increasingly interesting. Although the explicit linking between migration and development may have its roots a few decades earlier, the recent adoption of European Agenda on Migration has again drawn the attention to European Union’s increased effort to address underdevelopment as the root cause of migration. The Agenda itself is a significant project, bringing together more resources and funds to this cause than ever before, affecting a lot of third countries and therefore a lot of people. In addition, the 2015 ‘migration crisis’ and turn to ‘post-truth politics’, have further intensified the political situation surrounding migration.

Methodology

In this section I will elaborate on my methodological approach. The research-policy gap is defined as a discrepancy between academic research findings and public policies in the field of migration. In order to analyse the research-policy gap on the relationship between migration and development and _{to detect possible explanations for its existence, I will conduct an in-depth} qualitative content analysis of academic research and public policies. The object of the qualitative content analysis can be all sorts of recorded communication (Mayring, 2000). In this thesis, I will explore the existence _{of the research-policy gap on paper. The analysis will not} explore the processes of knowledge production or policymaking directly, but focuses on the output: the final written policy documents and academic publications. I will first analyse the existing base of academic research covering the relationship between migration and development, and the developmental drivers of migration in particular. This analysis will focus

(30)

on both the content as well as the dissemination of expert knowledge. It reviews the development of the research field and identifies the main trends and shared consensus. Hence, the analysis consists of a content analysis and quality assessment. The criteria for the assessment are, among others: theoretical development, empirical support, methodological approaches, quality of findings, the levels of consensus, accessibility and dissemination.

Secondly, I will analyse migration policies in detail, assessing to what extent expert knowledge is transferred and used. The policy analysis will initially be a manifest content analysis to detect to what extent these policies have adopted the ‘root cause approach’ and to identify ideas and elements that deviate from research findings. These findings will be categorized and structured. This categorization is done in an inductive manner. All documents will be read thoroughly in order to identify visible understandings and ideas supporting the policy approach. These understandings and ideas can be derived from multiple policy elements, such as the employed instruments, targets and expectations. Examples of the main categories are: identification of the drivers of migration, targeted migrant type, migration and development profiles of the targeted countries and regions, employed instruments and the expectations regarding the policy effect on development and the development effect on migration. The categories will be modified and refined throughout the process. However, since I also aim to explore the underlying rationales behind these policies, I will continue my analysis with a latent content analysis. This refers to the interpretation of content (Holsti, 1969). In this analysis, the focus is on discovering possible underlying meanings of the content (Babbie, 1992; Catanzaro, 1988; Morse and Field, 1995). Hence, through this method, I aim to detect, categorize and structure different rationales for the adoption of the ‘root cause approach’ and the possible utilizations of knowledge, based on the theories and conceptualisations of research-policy relations in my research framework. This concerns a more interpretative analysis to gain an in-depth understanding of the phenomenon.

Data selection and analysis

As previously discussed, the data for the content analysis consists of research and policy output, in the form of final written documentation. The academic research under consideration consists

(31)

of final documents, i.e. published, peer-reviewed academic works. A total of _{approximately 40} academic publications were thoroughly read and analysed. As for the public policies under analysis, the data selected comprises official policy documents, originating from the government entities who drafted and proposed the concerned policies, in this case the EU institutions. These are final documents, in the sense that they are all published and available to the public. This does not necessarily mean that the policies are finalized. Furthermore, since only official EU publications were selected, the views of other stakeholders in the form of political debates, consultation responses or other manifestations are not included. The document types selected for the analysis were, among others: communication documents, policy proposals, factsheets, progress and evaluation reports and strategic notes that were all publicly available online. A total of_{30 policy documents were thoroughly read and analysed.}

Limitations

This analysis has its limitations. It will be difficult to gain insights into the ‘messy realities’ and actual practices of research-policy relations. In order to fully explain how the discrepancies between academic research and public policies came about, extensive fieldwork is necessary into the practices of research-policy relations and dialogues, which falls outside the scope of this thesis. Hence, this thesis is moderate in its aims, as written documents alone will not sufficiently provide complete or clear-cut answers to my research questions. The method of content analysis is limited in its ability _{to find out what is occuring ‘behind closed doors’ in the policymaking} process. However, the in-depth analysis of both research and policies will enable us to retrieve some insights of research-policy dialogues. This thesis should therefore be regarded as a first endeavour into a more thorough analysis of this remarkable research-policy gap. Furthermore, with regard to the documents analysed, there is a dependency on the policy narratives of governmental agencies_{. The policy narratives and public rhetoric might disguise the hidden} objectives and agendas of policymakers. In addition, EU publications are also high-level documents, which implies that technical, detailed information of policy practices ‘on the ground’ is left out and often not made available to the public. The short review of progress and evaluation reports is not able to sufficiently overcome these limitations.

(32)

Social scientific relevance

The aim of this thesis project is to gain a better understanding into both the research and policy field of migration and development. It contributes to the literature on the migration-development debate as well as to the literature on research-policy gaps. As stated earlier, this thesis should be regarded as a first effort into a more thorough analysis of the research-policy relations and the understandings of policymakers concerning the migration-development relationship. The results can be helpful for both policymakers and researchers in the field, and inform researchers about what needs to be done to further explore the research-policy relations. The research-policy gap is quite remarkable as it exists for over thirty years, yet it has not received the sustained attention it deserves. This thesis aims to stimulate discussion and further academic research.

(33)

Chapter 3 The Migration-Development Nexus in Academic Research

As described earlier, the relationship between migration and development is studied by scholars from different academic disciplines and backgrounds. In general, the effects of development on migration have been particularly studied by migration scholars, as they are interested in migration processes, whereas the broader relationship between migration and development, including the effects of migration on development, is also largely covered by development studies. Hence, this analysis focuses on research carried out by migration scholars. However, as the issues are interconnected, some contributions overlap with development studies._{The analysis} covers the most relevant academic works. It reviews the development of the research field and identifies the main trends and shared consensus. The analysis includes both findings of migration studies as well as assessments of the research field by other prominent migration scholars. The first section sets out the theoretical development of the field, followed by an assessment of empirical findings. The third section discusses the strengths and weaknesses. Hence, the analysis consists of a content analysis and quality assessment. The criteria for the assessment are, among others: theoretical development, empirical support, methodological approaches, quality of findings, the levels of consensus, accessibility and dissemination. This chapter should provide answers to the first sub-question, whether there is sufficient expert knowledge on the migration-development relationship and thereby test if a lack of sufficient expert knowledge might explain the research-policy gap.

The ‘root causes approach’ to reducing migration seems to be broadly based on two assumptions, namely that (i) development aid can improve conditions in migrant sending countries and (ii) improved development or conditions cause fewer people to migrate (De Haas, 2006; 2007; Clemens and Postel, 2018). Thus, both the effect of aid on development as well as the effect of development on reducing migration should be positive in order for this approach to succeed. Whilst the effectiveness of development aid is also widely questioned (see e.g. Clemens

(34)

and Postel, 2018; Fratzke and Salant, 2018 in Chapter 1) most existing migration research has particularly focused on the effects of development on migration. Therefore, the primary focus here is on the determinants of migration and the effects of development on migration.

3.1 Migration theories

Functionalist theories

Since Ravenstein’s laws of migration, the perspective that people from low income to high income regions has remained dominant in migration studies (Ravenstein, 1885; 1889). Functionalist migration theories are based on ‘push-pull’ models and equilibrium assumptions, assuming an inversely proportional relationship between migration and income and opportunity differentials. Hence, migration is viewed as a function of spatial disequilibria, in line with functionalist social theory, which views society as a system with a tendency towards equilibrium (De Haas, 2010). According to De Haas, migration determinants research is still generally based ‘push-pull’ and gravity-based migration models, whilst they are, in his perspective, obsolete and theoretically void (De Haas, 2011). The push-pull models identify various factors that push and pull migrants from origin to destination and do not attribute any agency to migrants (De Haas, 2010; 2011).

Neoclassical migration theory explains migration by geographical differences in the supply and demand for labour. Accordingly, this theory views migrants as rational income-maximizing actors, who weigh the costs and benefits in their decisions to migrate. The early models were aimed at explaining internal rural-urban migration (Todaro, 1969; Harris and Todaro, 1970) but this was later modified to an international model (Borjas 1989). Although neoclassical migration models were largely focused on geographical imbalances in income levels, later adaptations included the costs of migration, to understand migrants’ selectivity and view migration as an investment (Bauer and Zimmermann 1998). Migration is then a form of investment in human capital. Later models still predict that rising average incomes in origin societies will tend to reduce migration (Clemens, 2014). These founding economic theories of migration seem to have several flaws. Although economic differentials play a major role in

Dissecting the gap : an analysis of the discrepancy between academic research and public policies on the migration-development nexus