!1
What are the Different Modes of Expert Groups Utilisation and How can They be Identified in Real-World Situations?
What Use of Expert Groups is Expected by the European Commission in Immigration Policy in Present Times?
Student: Sara Forato Student number: s2393247
Supervisor: Dr. Rik de Ruiter
Thesis submitted for the degree of Public Administration (MSc), specialisation in International and European Governance
March 11, 2020
Table of Contents
Abstract ……… 5
Introduction ……… 6
Theoretical Framework ………... 8
Literature Review ………... 8
The Entrance of Expertise in Policymaking ……….. 8
The Academic Study of Expertise ………. 10
The Supply Side Study of Expertise ………. 10
The Demand Side Study of Expertise ……….. 12
New Perspectives on Demand Side Studies ………. 14
The Limitations of Demand Side Studies ………. 15
The Contributions of this Thesis ……….. 15
Theory Building ……… 16
New Assumptions about Organisational Behaviour ……… 17
Strategies to Boost Legitimacy ………. 19
Organisational Features ……….. 20
Characteristics of Action Organisations ……….. 22
Characteristics of Political Organisations ……….. 22
Expertise Use in Action and Political Organisations ……….. 25
Definition of Instrumental and Legitimising Expertise Uses ………... 27
The Substantiating Strategy ………. 29
The Substantiating Use of Expertise ……… 30
The Consensus-Building Strategy ……… 33
The Consensus-Building Use of Expertise ……….. 34
Summary of the Chapter ……….. 36
Conclusion of the Chapter and Discussion of its Contributions ………. 38
First Contribution ……… 39
Second Contribution ……… 40
Methodology ……….. 41
Relevance of the Method ……….. 43
Within Case Analysis ……… 45
Document Analysis ……….. 48
!3
Indicators for Action and Political Organisations ……….. 51
Instrumental and Legitimising Expertise-Related Indicators ……….. 53
Indicators for the Substantiating Use of Expertise ……….. 55
Indicators for the Consensus-Building Use of Expertise ………. 56
Contributions in Terms of Operationalisation ………. 57
Analysis ………. 59
STEP 1 - Action or Political Organisation? ……….. 61
Indicators (1a) and (1p) ………... 62
Indicators (2ai) and (2aii) ……… 65
Indicator (2p) ………... 66
Indicator (4p) ……….. 67
Results - Political Organisation ……….. 69
STEP 2 - Legitimising Use of Expertise? ……… 69
Indicators (11pi), (11pii), (11piii), and (11pvi) ……… 70
Indicator (22pi) ……… 74
Indicators (44pi), (55pi), and (66pi) ……… 76
Results - Legitimising Use of Expertise ………... 77
STEP 3 - Adoption of a Substantiating Strategy? ……… 78
Indicator (8s) ……… 78
Results - Adoption of a Substantiating Strategy ……….. 80
STEP 4 - Substantiating Use of Expertise? ……….. 80
Indicator (11si) ……… 80
Results - Substantiating Strategy but no Use of Expertise ……….. 83
STEP 5 - Adoption of a Consensus-Building Strategy? ……….. 84
Indicators (1c), (9c), and (10c) ……… 84
Results - Adoption of a Consensus-Building Strategy ……….. 86
STEP 6 - Consensus-Building Use of Expertise? ………. 86
Indicators (11ci), (11cii), and (11ciii) ……….. 86
Indicators (44ci) and (66ci) ………. 87
Results - Consensus-Building Use of Expertise ………... 89
Outcomes of the Analysis ………. 89
Discussion ………. 90
Considerations on the Analytical Process ………... 93
Conclusion and Suggestions for Further Research ……….. 94
Conclusion ………. 95
!5 Abstract
In this thesis I will answer to my two research questions. The first one enquires in what way organisations can possibly use expertise and under which conditions such uses take place. In order to answer to it, I will engage in theory building. In particular, I will conceptualise four expertise utilisation modes - i.e. instrumental, legitimising, substantiating, and consensus-building. I will associate each of them with a number of variables determining when it is likely to be adopted. I will also develop indicators for every variable in order to ease their identification in real-world circumstances.
My second research question explores which utilisation mode the European Commission - Commission from now on - is expected to employ when dealing with immigration policy. The results of my analysis will indicate that the European body is likely to make a legitimising and a consensus-building use of knowledge. It is also expected to employ a substantiating strategy, although without involving experts. Additionally, the answer to this second research question will show that the theory building carried out in response to the first research question is meaningful and contributes to the literature on expertise use.
Introduction
Hereby, I will contribute to the demand side literature on expertise. This academic field is concerned with how the research produced by expert committees is employed by organisations. For a long time it was assumed that the only role of expert knowledge was improving organisational performance. However, the most recent literature on expertise utilisation has been based on the presumption that knowledge is used to improve organisational legitimacy. For example, an agency might want to improve its authority within its policy area by showing that it is able to meet its targets. Then, this agency might want to do so by gathering expertise about how to optimally fulfil its tasks. Alternatively, an administration may want to impose its policy preferences in order to be perceived as more authoritative. To show the superiority of its proposals, it may back them with expert knowledge.
Nonetheless, what are the expertise uses that agencies can possibly carry out? And under which conditions is an administration likely to employ one or the other? Scholars have tried to answer to these questions in many ways but their studies rarely built on each other, leading to a very fragmented demand side literature. Additionally, the only notable attempt to generate a comprehensive theory on expertise utilisation can be ascribed to Boswell (2012).
Given the relevance of this academic field, I will argue that more efforts must be directed towards the development of a demand side theory. As a matter of fact, Boswell’s account suffers from a number of shortcomings and needs refinement. My first research question targets precisely this necessity. More specifically it asks: what are the different modes of expert groups utilisation and how can they be identified in real-world situations? Hence, in response to it I will engage in theory building. More specifically, I will try to put forward an exhaustive collection of the possible expertise utilisation modes. I will identify four of them: instrumental, legitimising, substantiating, and consensus-building. Moreover, I will associate each of them with a set of variables indicating under which circumstances agencies are likely to adopt them. These variables will represent organisational and policy area characteristics that are identifiable in reality. I will also make them even more recognisable in real-world situations by associating them with indicators - i.e. through operationalisation. This will allow for the categorisation of administrations by the knowledge utilisation that they are likely to pursue.
To provide an example of how the new theoretical account that I will develop must be employed, I will apply it to a case. I will do so in my analysis, through which I will reply to my second research question. The latter asks: what use of expert groups is expected by the European Commission in immigration policy in present times? The application of the theory to a case will also clarify the
!7 strengths and weaknesses of the theory itself. In particular, it will be evident if it represents a step forward with respect to Boswell’s one. Providing an answer to this question will also shed light over a relevant and interesting case.
As it is clear, I will conduct a within case analysis. This means that I will gather rich, detailed, and contextual information. I will collect these data through document analysis. For example, I will rely on reports issued by the Commission, on the constitutional treaties of the EU, and on newspaper articles.
The results that I will attain - and the analytical process itself - will reveal that the theory developed in response to my first research question does ameliorate Boswell’s account. In other words, my contribution to the demand side study of expertise will prove meaningful. By answering to my second research question I will also uncover some weaknesses of my theoretical framework and I will indicate how they could be improved through further research. Last but not least, I will find out that the Commission is expected to use expertise with a legitimising and a consensus-building function when dealing with immigration policy. It is also likely to adopt a substantiating strategy that does not involve expert knowledge.
The thesis will unfold as follows. The theoretical framework is divided into two parts: literature review and theory building. In the former, I will better explain what kinds of research comprise the literature on expertise. More specifically, I will differentiate between supply and demand side research. I will also identify a gap in the latter. In the theory building part, I will try to fill this gap by starting to reply to my first research question. I will begin by considering the assumptions on which the most recent literature on knowledge utilisation is based. I will focus on Boswell’s research in particular. From there, I will start to conceptualise my four utilisation modes and I will put forward the variables associated with each of them. I will conclude this chapter by explaining how I believe to have enhanced Boswell’s theoretical account.
I will finish to answer to my first research question in the methodology chapter, where I will operationalise the variables formulated through my theory building attempt. As already mentioned, this will make them more easily identifiable in real-world situations. Here, I will also introduce my second research question and I will explain the choice of the case. In addition, I will illustrate how I will answer to it. Hence, I will describe how the analysis will unfold and which data collection and data analysis methods I will adopt.
In the analysis chapter, I will take into consideration the Commission and the policy field of immigration. By following the variables developed in the theoretical framework and the indicators
formulated in the methodology, I will try to determine which expertise use is expected by the Commission in this policy area. Once I will have obtained my results, I will juxtapose them and my analytical process with Boswell’s, so as to highlight the strengths of my theory compared to hers. Lastly, I will illustrate some of the weaknesses of my theoretical account and I will put forward suggestions for further research.
All these steps in the thesis will be accompanied by tables and figures that I have designed in order to make the text more understandable and easier to follow. Notably, each of the four utilisation modes is characterised by a different colour.
Theoretical Framework Literature Review
In this first part of the chapter, I will discuss the available literature on the role of expert groups in decision-making. More specifically, I will explain when and why this academic field gained relevance. Then, I will present the two main streams of research dealing with it. More explicitly, a supply or a demand side perspectives may be adopted to scrutinise the subject. Afterwards, I will identify a gap in the literature with regard to demand side research. I will focus on this analytical angle by explaining its development throughout time. I will then show its lacunae and argue for the reasons why they should be filled. In doing so, I will guide the discussion towards the topic of this thesis which, as a matter of fact, is policymakers’ demand for expertise and the factors shaping it. I would like to point out from the beginning that the trend to involve experts in politics encompasses every kind of administration (Boswell, 2012). This means that local, national, and transnational, as well as political and bureaucratic governmental bodies employ expert knowledge. The following theoretical discussion comprises all of them. Of course, various institutional and environmental features lead to a more or less intense involvement of experts. However, I will leave considerations about more specific circumstances impacting the use of expertise to the second part of the chapter. Therein, I will develop a theoretical framework allowing for the categorisation of administrations according to what use of experts they are expected to make.
The Entrance of Expertise in Policymaking
The diffuse presence of expert knowledge in politics has emerged as a research puzzle during the 1990s (Boswell, 2012, pp. 3-4; Radaelli, 1999, p. 761). This was the case because, as the complexity faced by decision-makers increased, the need for better-informed advisers became more impelling (Haas, 1992). The complicated and challenging nature of policy problems is a feature of
!9 politics that is present ever since. It is mainly due to the technical difficulties brought about by the phenomena that society must deal with nowadays (Haas, 1992; Rimkute & Haverland, 2015, p. 431). Scientific issues such as climate change and the rise of artificial intelligence, but also social challenges like population ageing, income polarisation, and democratic backsliding, are a source of unprecedented incertitude. Uncertainties encompass the preparation, interpretation, and implementation of legal texts. They also comprise the practical consequences of decisions and the very preferences on the matter held by politicians (Abbott, Levi-Faur, & Snidal, 2017, pp. 18, 20-21; Blom-Hansen & Brandsma, 2009, pp. 722-723; Häge, 2007, pp. 310-311; Radaelli, 1999, pp. 761-762). The impossibility of fully understanding issues prompts the emergence of various interpretations. These are, in the best-case scenario, inaccurate and likely to lead to unrealistic policy solutions and unforeseen outcomes. Also, diverse and incompatible problem frames impede communication and are conductive to deadlocks and suboptimal decision-making.
However, the complexity faced by policymakers has also an institutional character. In nowadays international arena, virtually every jurisdiction is involved to some extent in an extremely composite and layered architecture of multi-level governance (Eberlein & Newman, 2008, p. 27). Both private and public actors engage in interdependent decision-making which affects local, national, and transnational territorial aggregations (Piattoni, 2015, pp. 323, 325). However institutionalised these conglomerations of sovereign power are, the omnipresent shadow of network governance - coupled with the need to search informality - often causes unclarity in the decision-making process (Piattoni, 2015, pp. 325-327; Reh, Heritier, Bressanelli, & Koop, 2011).
In such a picture, authority seems to be lost in the incapacity to take well-informed and wise decisions and in the insecurity springing from reliance on a loose and seemingly uncontrollable governance structure. Hence decision-makers’ need to show to lay people that they have not lost control over the situation, and the necessity to adopt strategies actually making society governable. Evidence-based policymaking was the answer to these necessities (Boswell, 2012, p. 3; Gornitzka & Sverdrup, 2008, p. 728). Expert groups started to be involved at different stages of the decisional process. They can be composed by academics, representatives of national administrations, and representatives of various interest groups (Radaelli, 1999, p. 759; Robert, 2012, p. 430). More precisely, they are non-elected actors who possess authoritative knowledge on a specific field. The feature in virtue of which they became part of the policy process is their supposed ability to foster the achievement of optimal policy outcomes. They were seen as apolitical agents who base their decisions on neutral and objective knowledge (Robert, 2012, p. 427).
The Academic Study of Expertise
The literature on expertise in policymaking has approached and challenged this understanding of events from different angles. Before scrutinising these accounts, it is particularly intriguing to notice that after decades since the topic has started to arouse interest in the academic community, the discussion around it is still incoherent, poorly organised, and definitively unsystematic. Many of the contributions focus on the EU given the undoubtedly preeminent role that expertise plays in its administration (Blom-Hansen & Brandsma, 2009, pp. 719-721; Häge, 2007, p. 300; Radaelli, 1999, pp. 757-758; Robert, 2012, p. 426). In this context, three are the major outlets for expert knowledge, namely the European Council - Council from now on - working groups, comitology committees, and the Commission’s advisory groups (Rimkute & Haverland, 2015, pp. 433-434). Remarkably, each of them is usually analysed by a definite set of authors who end up specialising in the workings of the very committee more than on advancing the literature on expertise. Hence, the external validity of their publications is commonly low and there are almost no efforts to generalise findings. Moreover, there is no consistency among the theoretical approaches and the processes of conceptualisation adopted in the studies. This fact hinders comparability and, again, impedes generalisation. This is the case not only between analyses of different committees; the same applies to researches scrutinising the same expert group. In general, there is a sense of aimlessness with regard to the construction of theoretical frameworks whose scope goes beyond the mere ends of the single piece of academic work. A consequence of this modus operandi is that the process of theory building advances at a slow pace and in a scattered manner. Also, blindspots become difficult to detect.
For these reasons, hereby I will identify what I consider to be one of the most promising theoretical accounts on expertise. I will engage in theory building by merging it together with separate theoretical interpretations that can meaningfully contribute to the development of a more overarching theory. In this way, I will answer to the need for the selection and adaptation of sparse and overly contextualised approaches. Broader theoretical frameworks that organise fragmented knowledge allow for a more systematic theory application, assessment, and development. My first research question will be modelled on this need.
The Supply Side Study of Expertise
Academic works have approached the topic of expertise from two sides: supply and demand. The former scrutinises the dynamics that take place inside expert groups and how they impact decision-making. In other words, it determines which factors lead to certain resolutions within committees.
!11 Then, it explains the effect that these resolutions, when supplied to elected officials, have on the decision-making process overall.
Expectations on the behaviour of expert groups and on their relationship with policymakers are originated from the following theoretical reasonings. Firstly, according to theories on delegation of authority, decision-making power is bestowed upon more knowledgeable bodies by governmental authorities. This is done in order to bypass costly processes of information gathering and to overcome situations of legal, practical, or infra-actor relational uncertainty (Abbott, Levi-Faur, & Snidal, 2017, pp. 18, 20-21; Blom-Hansen & Brandsma, 2009, pp. 722-723; Eberlein & Newman, 2008, p. 27; Gornitzka & Sverdrup, 2008, p. 728; Häge, 2007, pp. 301-302; Hooghe & Marks, 2014, p. 307; Radaelli, 1999, pp. 761-762). Nonetheless, principals make sure to detain as much authority as possible over experts. Thus, only functions that are strictly necessary for carrying out the assigned task are conferred upon them (Christiansen & Dobbels, 2013, pp. 54-55; Radaelli, 1999, pp. 761-762). Consequently, theories of delegation predict that the greater the uncertainty faced by authorities, the larger the amount of power bestowed upon experts (Häge, 2007, pp. 310-311; Hooghe & Marks, 2014, pp. 311-313; Radaelli, 1999, pp. 762-764). Moreover, the smaller the gap between principals’ and experts’ preferences, the higher the level of authority enjoyed by the latter (Gornitzka & Sverdrup, 2008, p. 728; Häge, 2007, pp. 301-302). Measuring the extent of delegation is relevant in so far as it impacts the probability that decisional outcomes will mirror expert groups’ ideas.
Secondly, socialisation theories are usually employed to explain the way in which expertise might shape decision-making. Experts that are engaged at various administrative levels or in different venues are usually characterised by shared loyalties (Kaniok, 2016, pp. 884-885; Suvarierol, Busuioc, & Groenleer, 2013, p. 910). Socialisation implies that agents operating at the transnational level are likely to support more supranational stances (Blom-Hansen & Brandsma, 2009, p. 722; Suvarierol, Busuioc, & Groenleer, 2013). However, according to a more nuanced conception of the phenomenon, membership is only associated with the adoption of a certain behaviour that is accepted within the group. This understanding of socialisation entails that only negotiation and discussion styles are influenced, not the actual positions held by experts (Häge, 2007, p. 309; Häge, 2008, pp. 47-48).
This last point is a nice bridge towards the illustration of the third concept adopted in supply side studies: deliberation. Accordingly, various factors determine whether the members of a group will adopt a negotiation style leaning more towards deliberation or bargaining. The former leads to decisions that account for almost everyone’s preferences and usually allows for the achievement of
the common good. On the contrary, the latter produces outcomes that actualise powerful actors’ interests. If the group is homogeneous and characterised by individuals who share values, then deliberation is more likely to emerge. This reasoning stems from an understanding of socialisation as a factor able to change actors’ identity and preferences and make them more uniform (Blom-Hansen & Brandsma, 2009, p. 722). On the other hand, if socialisation means becoming acquainted with the rules governing interaction within a collection of individuals, then it will just strengthen the tendency to adopt a determinate communication style. In this case, higher socialisation might also lead to bargaining (Kaniok, 2016). Furthermore, technical complexity, regulatory instead of distributive issues, and low preference intensity are all expected to be conductive to deliberation (Blom-Hansen & Brandsma, 2009). Lastly, many studies focusing on negotiation processes have shown that if many actors with heterogeneous preferences are involved and a premium is placed upon long-term cooperation or avoiding deadlocks, then deliberative attitudes prevail (Fouilleux, de Maillard, & Smith, 2005, pp. 617-619; Juncos & Pomorska, 2007; Odell, 2009; Ostrom, 2010; Reh, Héritier, Bressanelli, & Koop, 2011; Risse & Kleine, 2010).
Lastly, the involvement of experts is usually associated with non-politicised issues (Radaelli, 1999). According to the actor constellations model developed by Genschel and Jachtenfuchs (2016), the very presence of mass publics in the discussion diminishes the weight carried by experts’ opinion (Dür, Bernhagen, & Marshall, 2015, pp. 957-958, 966-967). Moreover, politicisation is usually expected to favour the involvement of elected officials, thus hindering the prominence that expert groups would otherwise have enjoyed in decision-making. Indeed, political salience increases visibility and electoral returns (Häge, 2007, p. 311). Furthermore, when governing bodies want to avoid public scrutiny and discussion, decisional authority is conferred upon experts (Fabbrini & Puetter, 2016; Fouilleux, de Maillard, & Smith, 2005, pp. 611, 615-616; Radaelli, 1999). Again, understanding the dynamics connected to politicisation allows to estimate how much decisional outcomes will reflect expert groups’ views.
The Demand Side Study of Expertise
The second way in which expertise has been approached is by looking at its demand side. This body of literature examines how expert knowledge is used by governmental bodies instead of how it is produced. The rationale behind this focus is that whatever the content of decisions reached by expert committees and supplied to policymakers, with which ends they are going to be utilised is what actually matters in the policy process. In other words, expertise has a role in so far as it is used in some way. Its substance does not have an impact if it is not employed, or it does depending on
!13 the way in which it is utilised. Despite the relevance of this stream of research, more attention has been devoted to supply side analyses (e.g., Blom-Hansen & Brandsma, 2009; Christiansen & Dobbels, 2013; Fouilleux, de Maillard, & Smith, 2005; Häge, 2008; Juncos & Pomorska, 2007, Kaniok, 2016). My thesis aims at contributing to a more intense study of the demand perspective - as it has already been advocated (Metz, 2013, p. 269).
Research on expert knowledge utilisation has a long-standing tradition (Oh & Rich, 1996; Sabatier, 1987; Weiss, 1979, 1986, 1999). Scholars started off by focusing on a problem-solving notion of expertise, as I have illustrated at the beginning of this chapter. They identified functionalist reasons for policymakers’ demand for knowledge. Accordingly, they asked for more and higher quality information so as to take better decisions and improve policy outcomes. Consequently, political discussion would revolve more around efficiency and effectiveness than power politics and reputational benefits (Boswell, 2012, p. 4). This kind of instrumentalist understanding maintains that the generation of well-thought-out decisions is considered valuable per se - even though it also leads to a gain in output legitimacy (Gornitzka & Sverdrup, 2011; Gravier, 2013; Groeneveld & Van de Walle, 2010).
Nevertheless, the problem-solving notion soon started to be challenged by authors stressing that making sense of policymaking requires more nuanced characterisations of expertise utilisation. Weiss (1979) pioneered a broader understanding of the role of expert groups by identifying seven uses of knowledge (Rimkute & Haverland, 2015, p. 435). However, the presumption of her study was still that experts are valued mainly in light of their capacity to produce policy-relevant information that can be used to adjust policies (Boswell, 2012, p. 4).
It was not until decades later that new studies advanced the understanding of expertise utilisation by challenging and building upon its instrumentalist variant. Arguably, researchers realised the need to broaden their comprehension of demand side dynamics because of the related puzzle observable in the political realm. A gap exists between the policies that are adopted and the ones that are recommended by experts. In some cases, even administrative agencies owning in-house research centres seem to disregard experts’ findings. Politicisation has been employed to account for such behaviour. Accordingly, elected officials would choose electorally appealing solutions over evidence-based ones. Scarce resources to actualise policies prescribed by experts is another hypothesis used to shed light on the problem. Additionally, it has been argued that experts themselves are responsible for the limited use of their output (Boswell, 2012, pp. 5-6). In particular, they might not be able to present it in relevant venues, to convey its importance for real-world challenges, and to communicate it in an user-friendly manner. In other words, knowledge transfer
might be tricky (Boswell, 2012, pp. 32-34; Oliver & Cairney, 2019). However, political pressure, lack of organisational capacity, and perceived absence of relevant knowledge cannot explain the substantial trend of continued interest of governmental actors in expert groups and subsequent dismissal of their findings. Moreover, these theses continue to assume that, in so far as knowledge is used, it is done in virtue of its problem-solving property (Boswell, 2012, pp. 33-35). Therefore, the main question addressed remained whether expertise was going to be used at all and why (Boswell, 2012, p. 35).
New Perspectives on Demand Side Studies
It is clear that the discrepancy between recommended and actualised policies cannot be clarified through the instrumentalist account or through the alternative explanations that scholars have tried to put forward. Therefore, the need for theory building was evident. A limited number of scholars responded to this intellectual challenge and started thinking in terms of different purposes for which expertise could be employed (Boswell, 2008; Schrefler, 2010). They recognised the gap in the literature and began to look for alternative purposes fuelling governing bodies’ need for expert groups. They found answers in organisational and neo-institutionalist literature. What they identified as one of the major drivers for administrative action was the quest for legitimacy (Boswell, 2008, p. 473; Boswell, 2012, pp. 29-32, 40-47; Schrefler, 2010, pp. 311-313). Schrefler accounted for instrumental and symbolic - i.e. non-problem-solving - expertise utilisations. Moreover, she included the option of non-use (Schrefler, 2010). Although appreciable, her attempt is not as complete and as logically organised as Boswell’s (2008, 2012). The latter author attains a more parsimonious conceptualisation and is more rigorous in drawing from the academic fields relevant to the study of expert knowledge utilisation.
Other authors have acknowledged the importance and contributed to the development of non-instrumental accounts. In particular, Gornitzka and Sverdrup (2011) have put forward some symbolical explanations for knowledge utilisation that were not considered by Boswell (2008, 2012). They also connected these insights to specific features of expert groups. However, their findings were not generalised or linked to other demand side literature. In other words, they were valued only in light of the information they provided on a specific case. Besides, Metz (2013) seems to have built - consciously or not - on this new conception of expertise utilisation. More specifically, she theorised a symbolical use of expert groups that appears to be non-redundant when considering Boswell’s ones (2008, 2012). Nonetheless, this idea is not fully developed or operationalised. In addition, Metz’s discussion of instrumental and symbolical theories of knowledge utilisation is
!15 inexplicably selective and poorly framed. Subsequently, Rimkute and Haverland (2015) have adopted Boswell’s theoretical approach. Although I agree that it is the most complete and best theorised one, scholars should not disregard the relevance of other contributions such as the ones just mentioned. In this way, the two authors have complemented evidence but overlooked theory development.
The Limitations of Demand Side Studies
As already stated, all of these theoretical frameworks are conceptualised slightly differently in every piece of research and there is a proliferation of similar concepts. The lack of discussion about this tendency has prevented systematic empirical investigation and meaningful theory building.
Research continues to be produced on the instrumental use of knowledge and the factors enabling it (Cairney, Oliver, & Wellstead, 2016; Davies, Nutley, & Smith, 2000; Head, 2015). Instead, as it is evident from my overview, studies investigating other functions of expert groups are more limited in numbers despite their usefulness. In particular, the latter suffer from restricted efforts of theory building and methodical development allowing for a coherent application among researches. The problem-solving account enjoys a longer tradition of sustained interest and has the advantage of being based on rational-choice theory - which is characterised by clear and formalised assumptions. This is the reason why more intellectual work should be directed towards the theorisation of symbolical modes of expertise utilisation. In particular, research must focus on the development of an overarching framework accounting for every function that expert groups can have in policymaking. Moreover, a set of comparable conditions pointing to one or the other use must be constructed so as to improve applicability, reproducibility, and additional theory development. In other words, further conceptualisation and operationalisation must be carried out.
The Contributions of this Thesis
Hereby, I will move some first steps in undertaking this task. First of all, I will improve the understanding of symbolical accounts. In particular, I will add to Boswell’s theory insights that emerged in later literature. In other words, I will supplement it with a new expertise utilisation mode. In this way, I will contribute to the development of an overarching theory comprising problem-solving and non-instrumental perspectives. Secondly, I will enhance the conceptualisation and operationalisation of the factors pointing to the various utilisation modes.
These two contributions answer to the first of my research questions: what are the different modes of expert groups utilisation and how can they be identified in real-world situations? They refine the
theoretical and methodological tools associated with the most promising demand side theory on expertise. More practically speaking, this improved version of the theory will allow for a better and more systematic categorisation of administrations according to the knowledge use that they are expected to make. In other words, given an agency, the theory will anticipate which use of expertise it will likely pursue. This will be done by analysing some of the features of the agency according to a definite collection of characteristics.
I will then apply this newly formed demand side theory to a case. In this way, I will show how it must be employed and I will understand its strengths and limits. Additionally, the chosen case is relevant in and of itself. This step answers to the second research question: what use of expert groups is expected by the European Commission in immigration policy in present times?
Hopefully, my theory building attempt will lead to a more complete and more reusable demand side theory. Furthermore, I hope to manifest the need for similar endeavours and to encourage other researchers to give their contribution in line with the suggestions that I will put forward throughout this paper.
Theory Building
In this second part of the chapter, I will try to answer to the need for further development of a demand side theory of expertise. While in the methodology chapter I will focus on the operationalisation, here I will concentrate on the conceptualisation of variables and of an additional expertise use. Therefore, I will begin to target the first research question.
I will mainly use Boswell’s theoretical account (2008, 2012) as foundations for my theory building. Indeed, as already stated and as asserted by Rimkute and Haverland (2015) as well, her classification of expert knowledge utilisations is more logically rigorous, parsimonious, and yet comprehensive than other authors’ ones (e.g., Schrefler, 2010).
The aim of a demand side theory is to determine in which way administrations are expected to use expert knowledge. In order to do so, Boswell relies on the analysis of some institutional features. I will do so as well. In particular, I will outline sets of variables associated with each utilisation mode.
In order to build on existing accounts, I must take into considerations the assumptions on which they are based. Hence, I will start by describing how scholars borrowed from organisational and
!17 neo-institutionalist literature to construct a theoretical account of expertise utilisation that could overcome the problems associated with functionalist theories. According to their findings, I will argue why achieving legitimacy must be considered an organisational objective separate from performance-oriented goals.
Then, I will describe how the pursuit of legitimacy affects administrative action. In particular, I will outline two different strategies to achieve it and I will explain which organisational features are associated with each. Hence, I will introduce the concepts of action and political organisations. Action organisations employ an output-based legitimising strategy. On the other hand, political organisations rely on a discursive one.
Only after these theoretical considerations I will introduce expert knowledge in the discussion. In particular, I will put forward additional variables that determine if, within the scope of these two strategies, expertise is likely to be employed. Indeed, both courses of action can be pursued without implicating experts.
As a consequence, I will explain that the set of organisational features together with the factors associated with the use of expert knowledge determine when agencies are expected to make an instrumental or a legitimising use of expertise. Thereafter, I will follow the same steps in the conceptualisation of the substantiating and the consensus-building utilisation modes.
New Assumptions about Organisational Behaviour
As mentioned in the first part of the chapter, both Schrefler and Boswell started by looking for the goals motivating administrations’ behaviour. According to a rationalist account, they are believed to be concerned only by their policy output (Schrefler, 2010, p. 311). However, this explanation does not describe real-world phenomena satisfactorily (Boswell, 2012, pp. 32-36). As maintained by organisational literature, administrations are first and foremost preoccupied with securing legitimacy (Boswell, 2008, p. 473).
Even though almost no author openly covers them, instrumentalism is grounded on some deeply-rooted assumptions on organisational behaviour. The first one is that administrations work to fulfil performance-oriented objectives (Boswell, 2012, pp. 30-31). They do not judge their mandated duties, they only deliver on them. Second, agencies are unified actors, meaning that their output can be attributed to the sum of the activities performed by the individuals composing them (Boswell, 2012, p. 31). To sum up, bureaucracies are not autonomous actors so they cannot be ascribed own preferences. Moreover, the rational nature of cumulated individual acts leaves group dynamics and the problem of collective action out of the picture.
On the other hand, neo-institutionalist theories look at organisations through lenses other than the rational-choice one. More specifically, they maintain that bureaucracies are characterised by their own interests and that they employ various strategies to pursue them (Boswell, 2012, pp. 40-41; March & Olsen, 1984, pp. 738-739). With regard to the first claim, organisations are not only concerned with meeting targets and providing determinate outputs. Instead, they continuously try to reach a balance between the resources devoted to action and to politics (Boswell, 2012, p. 251; Schrefler, 2010, pp. 312-313). This is an insight that both Boswell and Schrefler built upon in their works. They derived it from Brunsson (1985, 1989) who, describing the structure of organisations from a non-functionalist point of view, highlighted the symbolical aspect of their actions (Schrefler, 2010, pp. 312-313).
Accordingly, the main driver of administrative behaviour is securing legitimacy (Boswell, 2012, p. 41). In order to do so, agencies need to handle both internal and external pressures (Schrefler, 2010, p. 313). Therefore, it is possible to distinguish between internal and external legitimacy.
The former has to do with the management of members. Of course, organisations are preoccupied with ensuring their own existence, if nothing else because it is a necessary condition for the achievement of mandated objectives. In order to do so, they need to be functional, which means that members must be able to work. To allow for this, agencies allocate resources to the creation and perpetration of an organisational culture. This process is often unconscious but it crucially leads to the generation of a cognitive framework that members use to make sense of the environment in which they operate. They come to share norms, beliefs, and routines that reduce the overload of information processing and help rationalising events and avoiding emotional distress. Then, individual action is not based on rational and efficiency-driven judgements, but on expectations and obligations.
Therefore, organisations become interested in reproducing their culture. This task becomes a goal in its own right and it is therefore perceived as different from improving output. It is important to notice that being deep-seated and unquestioned, organisational norms and practices are difficult to completely identify and change. Consequently, even when they hinder functionality, a logic of appropriateness is likely to keep driving administrations’ and their members’ behaviour (Boswell, 2012, pp. 41-43; DiMaggio & Powell, 1983; March & Olsen, 1984, pp. 740-741, 744).
External legitimacy has to do with organisations’ strife to meet expectations from the environment they inhabit. This external domain comprises the actors that play a role in the policy context within
!19 which agencies have influence (Schrefler, 2010, p. 313). Expectations might derive from stakeholders, competing or complementary administrations, costumers, and political personalities. It is in the interest of organisations to obtain and protect external legitimacy because their very existence is dependent upon the resources that other entities provide them with. The latter are willing to finance and support administrations if they fulfil their functions. Nonetheless, supplying certain outputs is not enough to show that they are delivering on what is expected from them. Instead, organisations are required to comply with a set of behavioural standards and routines. Hence, they need to invest resources into openly displaying that they are functional so as to attract the means necessary for their survival (Boswell, 2012, p. 43).
Because it is so fundamental and because it is dependent on something else than just meeting objectives, securing external legitimacy becomes an organisational goal in and of itself.
Therefore, the contributions illustrated in this paragraph maintain that organisational action is not just based on performance-oriented considerations. Instead, agencies make great effort to accomplish other goals. Namely, they pursue internal and external legitimacy. I will now describe how administrations act in order to boost their legitimacy.
Strategies to Boost Legitimacy
The second contribution of neo-institutionalism concerns precisely the ways in which organisations go about gaining legitimacy. Since agencies are committed to the achievement of internal and external legitimacy, organisational strategies are a by-product of their internal assets and external environment (Schrefler, 2010, p. 313).
As already mentioned, agencies are supposed to derive legitimacy from their output (Boswell, 2012, p. 43). If an agency delivers on determinate objectives, members should feel secure about its ability to overcome challenges and external actors should consider it efficient. However, the achievement of mandated goals might be difficult to identify and specific courses of action that an administration undertakes to improve its output might not be perceived as optimal or efficacious. Therefore, the pursuit of both internal and external legitimacy requires the employment of additional strategies. It is fundamental to realise that organisations need to demonstrate that they act in an appropriate way (Boswell, 2012, pp. 43-45). That is, in a manner that is considered by members and actors from the outside as the proper one to improve performance. In other words, they must act according to expectations. Such rules of conduct have nothing to do with enhancing functionality, they just show commitment.
Consequently, organisations engage in what can be considered perception management. They do so through their talk and decisions. The former has to do with employing a certain rhetoric and rationalising conduct. Given values and activities are exposed and an appropriate communication style is adopted. The latter involves making sure to be seen as reacting to problems that are perceived as falling within the scope of an agency. It also entails framing matters as requiring action or not. Importantly, the actual addressing of issues does not have an impact here (Boswell, 2012, pp. 43-44).
Thus, until now I have argued that organisations pursue product-oriented and legitimacy-based goals. The latter are achieved both through output-based and discursive techniques. Notwithstanding that legitimising strategies are employed by organisations, which ones are implemented and to what extent is dependent on a number of factors. Noticeably, the adoption of strategies founded on persuasion is connected with the opportunity to derive legitimacy from output. Agencies that can obtain reasonable levels of legitimacy through performance are less reliant on their talk and decisions. In the next paragraphs, I will lay out the variables determining which strategies agencies choose in order to respond to internal and external pressures.
Organisational Features
Organisations’ choices about the course of action to implement vary according to their mandate, internal capacity, and external context (Schrefler, 2010, p. 313). Therefore, by analysing these elements, expectations can be derived about organisational behaviour (Boswell, 2012, p. 46). I will now put forward a more detailed collection of the organisational features that need to be taken into consideration when trying to determine what courses of action administrations are likely to follow. The first element rests upon agencies’ mandate. External actors’ expectations will be derived - among other things - from how, formally, a given administration is supposed to behave. Then, (1) sticking to its mandate as a way to respond to monitoring might allow an administration to gain legitimacy (Schrefler, 2010, p. 312).
Secondly, two external forces holding expectations regarding proper behaviour are the public - i.e. targets - and political principals (Boswell, 2012, pp. 49-50). As already stated, organisations are dependent on the latter in order to obtain vital resources. Thus, (2) they promptly respond to the expectations of whom they hold themselves accountable to in the political system (Boswell, 2012, pp. 50-51, 67). To be sure, this variable also implies that principals are aware of the changes operated by the organisation (Boswell, 2012, p. 54). In other words, the chain of accountability
!21 must be operational. With regard to the former, it could be argued that if an agency has close interactions with its costumers and if it experiences their feedback firsthand, then it is likely to adjust its actions according to their demands. However, public claims about desirable forms of intervention must be filtered through political representatives in order to be heard. Some officials must take up citizens’ interests in order for changes to take place. Therefore, administrations are rarely required to directly respond to a shift in public expectations (Boswell, 2012, pp. 50-51; Hooghe & Marks, 2008, pp. 5, 16-18). Hence, (3) politicisation induces behavioural change if ridden by the political leadership. However, this variable is a particular case of agencies’ dependence on principals. Consequently, I will not treat it as a separate element in the typology. Thirdly, the external environment confronted by administrations has an impact that is independent from the preferences of the agents inhabiting it. More specifically, uncertainty might prevent organisations from understanding the situation they face and from building a legitimising strategy that responds to it. Thus, (4) when confronted with incertitude, they resort to safe actions and to the endorsement of good practice (Boswell, 2012, p. 68; DiMaggio & Powell, 1983, p. 151).
Fourthly, organisational responses to the internal environment must be understood in light of the organisational culture. As already mentioned, shifts in members’ cognitive frames are slow and difficult to implement. Therefore, (5) an administration will try to maintain coherence vis-à-vis normalised practices and shared values and beliefs. Challenging them is likely to make members unable to operate, thus depriving the organisation of one of its main resources.
These features characterise every administration. Nevertheless, different agencies will face diverse kinds of internal and external pressures and their mandate will require distinct attitudes. In particular, the literature distinguishes between action and political organisations. The two types are characterised by diverse mandates, internal structure, and exogenous environment.
Then, these distinct traits are linked to which legitimising strategies are employed. Action organisations are prone to derive legitimacy from their output, whereas political ones are likely to employ talk and decisions (Boswell, 2012, p. 47).
Table 3.1 - Typology of organisational characteristics 1 Mandate
2 Principals’ expectations 4 Uncertainty
5 Organisational culture
Table 3.2 - Organisational characteristics
Action organisations Political organisations
1a The mandate focuses on results that must be
delivered. 1p The mandate focuses on prescribing structural adjustments. 2a Resources invested in understanding the
outcomes desired by principals. 2p Homogeneity among organisations or between organisations and principals. 4p Homogeneity between related organisations
under conditions of uncertainty. 6a Ability to pursue the desired outcome.
7a Accreditation of results.
Table 3.3 - Instrumental and legitimising expertise utilisation modes
Instrumental Legitimising
Strategy-related features Action organisation - i.e. pursues an
output-based legitimising strategy. Political organisation - i.e. pursues a discursive legitimising strategy. Expertise-related features
11a Use of descriptive knowledge. 11p Expertise is considered authoritative in the policy community.
22a Use of explanatory knowledge. 22p Research in uncertain and potentially harmful policy areas when the organisation is
accountable. 33a Use of predictive knowledge.
In the next paragraphs I will explain which characteristics - derived from the above typology - are associated with the two organisational types. In this way, I will be able to illustrate which factors point to the implementation of one or the other legitimising technique.
Characteristics of Action Organisations
Agencies can be classified as action ones if their internal and external environments expect to witness output adjustments and if they are capable of affecting their output (Boswell, 2012, p. 49). Let me consider the first condition. Here the organisational mandate is relevant since (1a) if delegation of tasks took place because determinate outcomes needed to be delivered, then an agency can be labelled an action one. Moreover, (2a) action organisations invest efforts into carefully understanding which interventions their political principals want them to pursue. Consequently, close interactions and frequent contact must be observed. Conditions of uncertainty are not expected to have a big impact on the conduct of action organisations since they seek good practice by default - i.e. in normal conditions as well as in ambiguous ones. As regards features of organisational culture ascribable to action organisations, the literature makes no mention. Scholars usually argue that this is the case because of excessive variability (Boswell, 2012, p. 46). Organisational culture is also hardly observable by non-members. Thus, when possible, the impact of the fifth variable must be evaluated on a case-by-case basis.
The second feature associated with being an action organisations is that agencies must perceive that they are capable of inducing the change that they are asked for. The fulfilment of this criterium is also dependent on organisational narratives and beliefs about how much action will be recognised by principals. Thus, both (6a) ability and (7a) accreditation are fundamental (Boswell, 2012, pp. 53-54). This condition is somehow connected to the accountability requirement leading to reaction to political principals which, however, does not entail ability. Therefore, this criterium must be considered a specificity of action organisations.
Characteristics of Political Organisations
On the other hand, political organisations respond to internal and external pressures by adjusting their rules, structure, and rhetoric. This legitimising mode is linked to the concept of isomorphism. The latter illustrates the reasons why administrations are likely to adopt appropriate behavioural standards. Hence, it can guide towards understanding the features of political organisations. More specifically, isomorphism is the process through which a unit in a population comes to act and look like other units facing the same environment (DiMaggio & Powell, 1983, p. 149). In light of the
!23 present discussion, two reasons can be identified for this to take place. The third process described by DiMaggio and Powell bears low significance in multilevel governance systems because it is founded on homogeneity of actors (DiMaggio & Powell, 1983, pp. 152-154).
Coercive isomorphism results from pressures exerted on organisations. Such exhortations to adopt determinate norms and structures can be formal and direct or subtle and veiled. In the former case, requests might take the form of mandate (DiMaggio & Powell, 1983, p. 150). (1p) Political principals are prone in prescribing adjustments of standards in political organisations. Such changes are often ceremonial and founded on value-laden motivations (DiMaggio & Powell, 1983, p. 150; Finnemore & Sikkink, 1998, p. 895).
In the latter instance, organisations try to resemble authoritative models in a given field. This may imply copying from more powerful rival agencies, abiding by principals’ desires concerning institutional setups, or developing the same practices and structures characterising principals. In the first case, it is assumed that more authority and trustworthiness will be ascribed to an administration embracing rules commonly seen as reliable. Moreover, group dynamics of praising conformity and ridiculing deviation are soft mechanisms leading to norm adoption. For what concerns the second case, administrations win legitimacy vis-à-vis their principals when following their preferences. The rationale behind the third instance is that interaction is facilitated. Hence, the costs of collaboration wear down and the probability to win economic or political support increases. In all cases, changes are implemented so as to gain leverage and legitimacy in the eyes of principals. In fact, action is not mandated but modelled on the mental frames of whomever can direct the allocation of resources (Boswell, 2012, p. 67; DiMaggio & Powell, 1983, pp. 150-151, 154-156; Finnemore & Sikkink, 1998, pp. 895, 902-904). (2p) “As a result, organizations are increasingly homogeneous within given domains and increasingly organized around rituals of conformity to” patron organisations or institutions legitimised by decision-makers (DiMaggio & Powell, 1983, p. 150).
On the other hand, mimetic isomorphism does not stem from external actors but from the uncertainty pervading organisations’ environments. Unsureness may be high about the expectations and desires of exogenous agents, the consequences of action, the publicity of action, the abilities of the administration itself, and the internal endorsement of the organisational ideology. Under any of these conditions, political organisations cannot rely on output and their legitimising goals are obscure. Therefore, they adopt a mimic conduct, which they may consider the only viable and cheap solution. Here, the preferences of principals do not determine organisational behaviour. On the contrary, (4p) ambiguity causes organisations to model themselves on agencies they have contact with. Since they adjust their formal rules and informal norms also in conditions of low uncertainty,
they resort to the same course of action when incertitude arises. They seek good practice but, in the impossibility of identifying one, they copy a rather random administrative structure (DiMaggio & Powell, 1983, pp. 151-152, 154-156). “The ubiquity of certain kinds of structural arrangements can more likely be credited to the universality of mimetic processes than to any concrete evidence that the adopted models enhance efficiency” (DiMaggio & Powell, 1983, p. 152).
In this paragraph, I have laid down the features leading to legitimising practices based on decisions and talk. The latter characterise political organisations. Beforehand, I have illustrated the elements that are necessary to identify action organisations - which rely on output to legitimise themselves. Hence, I have formulated two sets of organisational features, each leading to a distinct legitimising strategy. Table 3.2 illustrates them. As the discussion above has made clear, I have generated these variables following the ideas of DiMaggio and Powell (1983), Finnemore and Sikkink (1998), and Boswell (2012). Nonetheless, their accounts represent just the roots of my work. Indeed, I have extensively reorganised and developed them. I have also adapted them to my typology of organisational characteristics (see table 3.1). Therefore, the information enclosed in table 3.2 is predominantly my own.
I will now - for the first time in this second part of the chapter - introduce the concept of expertise. Indeed, now it is important to determine when administrations use knowledge produced by experts as part of their legitimising strategy. As a matter of fact, both courses of action based on manipulating output and on controlling decisions and talk might employ expertise. Therefore, the use of expert knowledge must be understood as one of the techniques adopted within a more overarching legitimising strategy founded either on product management or discursive approaches.
1 Mandate
2 Principals’ expectations 4 Uncertainty
5 Organisational culture
Table 3.2 - Organisational characteristics
Action organisations Political organisations
1a The mandate focuses on results that must be
delivered. 1p The mandate focuses on prescribing structural adjustments. 2a Resources invested in understanding the
outcomes desired by principals. 2p Homogeneity among organisations, between organisations and principals’ desires, or between organisations and principals. 4p Homogeneity between related organisations
under conditions of uncertainty. 6a Ability to pursue the desired outcome.
7a Accreditation of results.
Table 3.3 - Instrumental and legitimising expertise utilisation modes
Instrumental Legitimising
Strategy-related features Action organisation - i.e. pursues an
output-based legitimising strategy. Political organisation - i.e. pursues a discursive legitimising strategy. Expertise-related features
11a Use of descriptive knowledge. 11p Expertise is considered authoritative in the policy community.
22a Use of explanatory knowledge. 22p Research in uncertain and potentially harmful policy areas when the organisation is
accountable. 33a Use of predictive knowledge.
!25 I will now explore the additional conditions under which action and political organisations turn to expertise.
Expertise Use in Action and Political Organisations
“Even where organizations do seek to derive legitimacy through adjusting their output, they will not necessarily feel compelled to draw on expert knowledge to help achieve the desired impacts” (Boswell, 2012, p. 55). Nonetheless, they will do so if new information is considered valuable in order to make functional changes and enhance output.
In this framework, the variables pointing to the quest for efficiency-improving knowledge are three. The first one is (11a) an organisation’s search for descriptive knowledge. This kind of expertise offers the illustration of phenomena faced by the agency. It sheds light on what is happening and on the scale of the events. Without it, administrations would not know what they are targeting, which would hamper action. Secondly, (22a) action organisations might need explanatory knowledge since it clarifies the causal links between a problem and its determinants. This expertise helps to understand how potential organisational interventions would impact the situation. This means that, without such knowledge, the administration would be incapable of determining which output adjustments are needed and whether the changes would meet the two conditions associated with action organisations - i.e. expected output adaptation and ability and accreditation. Importantly, explanatory research is usually structured so as to highlight the causal mechanisms that are susceptible to an administration’s action. Thirdly, (33a) predictive knowledge may be employed in order to forecast future developments of a particular issue area. It is based on the generalisation of assumed causal relations and it makes it possible to account for upcoming challenges and design future-proof and long-lasting policy responses. This type of knowledge is particularly precious if the problem addressed by an agency is ever-changing (Boswell, 2012, pp. 55-58).
With regard to political organisations, there are conditions under which they are prone to depict their talk and decisions as based on expert knowledge. Their attitude towards expertise utilisation refers back to isomorphism. More specifically, it has to do with which organisational structures and rules are chosen to be copied (Boswell, 2012, p. 69). Here, I will describe which factors entail the adoption of expert knowledge. Accordingly, organisational rhetoric must involve the rational justification of action and administrations must show that their decisions comply with experts’ advices.
Recalling that discursive strategies are adopted to follow the expectations of internal and external actors in case of coercive isomorphism, (11p) organisations will turn to expert groups for input when operating in a policy community in which expertise is deemed essential. This knowledge will not be useful in so far as it informs policymaking. On the contrary, its mere displayed presence fulfils a legitimising function (Boswell, 2012, pp. 70-72).
The second factor refers back to mimetic isomorphism and it is founded on the uncertainty pervading the external context. The dimension of incertitude that is especially relevant here is the impossibility to determine the societal impacts associated with organisational decisions. Administrations worry particularly about the unknown consequences of action in policy contexts in which the potential for harmful societal effects is high (Boswell, 2012, p. 71). Indeed, they fear assuming responsibility for possibly dangerous or undesirable decisions if future outcomes are unclear. This is the case because their legitimacy may suffer a plunge. In this situation, agencies draw on expert knowledge in order to shield themselves from legitimacy losses. In other words, they shift accountability to expert groups by depicting organisational decisions as following actions recommended by them. In this way, they will not be blamed and condemned to a decrease in legitimacy in case unwanted societal consequences will arise. Therefore, (22p) accountable administrations “have an interest in showing that they have carefully weighted up available evidence, through rational decision making procedures” by gathering and undertaking research in policy areas defined by uncertain and potentially harmful societal effects (Boswell, 2012, p. 71).
To summarise, organisations seek legitimisation through action adjustment if the agents populating their environment expect to witness output changes and if such modifications can be actualised by the administration in a perceivable way. Consequently, expert knowledge is employed as a tool to increase legitimacy if it has a descriptive (11a), explanatory (22a), or predictive (33a) function. Moreover, the concept of isomorphism explains the reasons pushing political organisations towards the adoption of a discursive legitimising strategy. Nonetheless, additional criteria must be added to investigate when this kind of agencies integrate the use of expert knowledge in their talk and decisions. This can take place as a form of coercive isomorphism if a decision-making style based on research is more or less imposed by actors populating the policy context (11p). Expertise may be used as a source of legitimation also as a mimetic response to uncertainty about societal outcomes and to responsibility for such unforeseeable effects (22p).
!27 Definition of Instrumental and Legitimising Expertise Uses
These different ways of using expert groups represent the knowledge utilisation modes that are studied in this thesis. The theoretical discussion above illustrates how neo-institutionalist and administrative theories have offered a more complete account of organisational goals. In particular, they have argued that legitimisation must be added to the performance-oriented objectives pursued by administrations. Then, scholars have focused on explaining how legitimacy is chased and which organisational and environmental characteristics determine the adoption of determinate strategies in order to do so. In particular, researchers interested in expertise have tried to understand which additional circumstances lead to the use of expert groups in the context of these legitimising strategies.
Therefore, in order to classify organisations according to the expertise use that they are expected to make, two sets of variables must be considered. Firstly, it is necessary to look at organisational features; they are needed to determine which legitimising strategy is pursued. Then, additional expertise-related variables must be checked to see whether experts are involved in the strategy. Accordingly, action and political organisations have different features and therefore diverse ways to achieve legitimacy. Moreover, whether expert knowledge is used to gain output-based or discursive legitimation also depends on specific features of the situation faced by agencies.
Therefore, constructing categories composed of strategy-related and expertise-related variables is how this theoretical framework tries to answer to the conceptualisation part of the first research question. This has never been done by demand side researchers. Then, each category points towards one mode of expertise utilisation.
Now I will better define and name the expertise uses and link them to the discussion above, starting with the instrumental use of knowledge. This utilisation mode is adopted by action organisations (Boswell, 2012, p. 49). The general conditions characterising them are the need for performance adjustments to please other actors and having the capacity to manipulate output and being accountable for such change. Multiple variables are implied by these criteria, for example the direct imposition by principals of targets to be met (1a). Most of these variables are derived by the organisational characteristics identified by organisational theory as the drivers of agencies’ action: their mandate (1), external pressures (2), endemic uncertainty (4), and organisational culture (5). Additional ones must be introduced to account for the difference between action and political organisations. Ability (6a) and accreditation (7a) must be considered in the definition of the former.
