COI reports: objective truth or political tool? The influence of domestic politics on the production of Country of Origin Information reports

(1)

COI reports: objective truth or political tool?

The influence of domestic politics on the production of Country of

Origin Information reports

University of Amsterdam

MSc Political Science

Migration, Territory and Power: Governing and Controlling Human Mobility

Student: Yvonne Rijnja

Supervisor: Dr. Darshan Vigneswaran

Date: 03-07-2020

(2)

1 List of abbreviations

ACCORD Austrian Centre for Country of Origin and Asylum Research and Documentation

COI Country of Origin Information

DAF/CAB Directie Sub-Sahara Afrika, Cluster Ambtsberichten [Sub-Saharan Africa Department, Cluster for COI reports]

DEMIG Determinants of International Migration

DFAT Department of Foreign Affairs and Trade [Australia]

EASO European Asylum Support Office

ECOI European Country of Origin Information Network

FGM Female Genital Mutilation

IARLJ International Association of Refugee Law Judges

ICMPD International Centre for Migration Policy Development

IDP Internally Displaced Persons

IMPALA International Migration Policy And Law Analysis

IMPIC Immigration Policies in Comparison

IPOB The Indigenous People of Biafra

LGBTI Lesbian, gay, bisexual, transgender and intersex

MASSOB Movement for the Actualization of the Sovereign State of Biafra

MEND Movement for the Emancipation of the Niger Delta

NGO Non-governmental organization

(5)

2 Introduction

Worldwide migration has grown rapidly in recent years, including the movement of refugees (United Nations, Department of Economic and Social Affairs, Population Division, 2017). The number of asylum seekers has grown to over 4 million worldwide, and the number of refugees to over 26 million (UNHCR, 2020). These growing numbers have led to more asylum applications and receiving countries are experiencing more pressure on their status determination departments. Receiving countries are not always able or willing to accept all the individuals that seek protection in their countries, but are obliged to protect the individuals that would otherwise experience persecution in their country of origin. Therefore, it is highly important that the countries that receive asylum claims thoroughly assess who is in need of protection.

However, the growing number of refugees and asylum seekers, as well as more complex flight routes and more diverse countries of origin, have made this determination procedure more complicated (Engelmann, 2015, p. 110). Some scholars even introduced terms such as ‘refugee roulette’ to highlight the variation in outcomes of similar asylum applications (Tsangarides, 2010; Ramji-Nogales, Schoenholtz, Schrag, 2007). However, in the status determination process, decision-makers decide on lives of individuals by assessing their risks of persecution in their countries of origin, and the consequences of a wrong assessment of an applicant’s risk could therefore be significant. If an asylum seeker is wrongfully forced to return to their country of origin, they could be at risk of persecution or even death (Tsangarides, 2010, p. 8). Therefore, it is crucial that the status determination process is objective, well considered and thus fair for the applicants, despite the growing pressure on the asylum decision-makers.

To make the process fairer and more efficient, almost all (Western) countries produce Country of Origin Information (COI) reports1. These reports consist information about the

situation in the countries of origin of the applicant and have grown in importance since the early 2000s (ACCORD & Austrian Red Cross, 2013, p. 12; Good, 2015, p. 123; Van der Kist, Dijstelbloem & de Goede, 2019, p. 69; Gyulai, 2007, p. 9). Governments have composed so-called COI units that collect COI from different sources and compare this information to eventually produce one report. The reports are essential for the status determination process not

1 I will use the term COI reports in this thesis for the sake of simplicity. In practice, these COI can not only be presented in the form of reports, but can also be ‘… fact sheets, responses to specific queries and documentation packages’ (Aggarwal & Floridi, 2020, p. 1).

(6)

only because it saves a lot of time for asylum decision-makers because they don’t have to search for background information on the country of origin and compare all the available sources, which leaves them with more time to delve deeper into an applicant's personal circumstances and thus better assess their future risks of persecution. It also should lead to a fairer status determination process because the country information is (or is claimed to be) well-assessed by the COI units that collect and compare COI from different sources and bring all the information together in one report. Because of its importance, it is crucial that the COI reports contain neutral and accurate descriptions of the security situation in a country of origin. However, reports on similar countries of origin can differ between reporting countries (Engelmann, 2015, p. 115). This raises questions because one would expect that if countries made assessments of the security situation of a country of origin at a similar time, they would all come to the same conclusions.

One possible explanation for the differences between COI reports could be political influence. In 2014, for example, the Danish Ministry of Justice stated that on basis of new country of origin information from fact finding missions, the refugee status of Eritreans who had left the country illegally or fled service in the army would be suspended (Van der Kist et al., 2019, p. 68). Strikingly, during that time, the Danish government received criticism from their society about the growing number of asylum seekers. Several human rights organizations accused the Danish government of strategically using COI to substantiate and legitimize a more restrictive asylum and refugee policy (Rosset & Liodden, 2015, p. 26). This example suggests that there is a relationship between political preferences and COI reports that are produced by governmental COI units. Van der Kist et al. (2019) and Rosset and Liodden (2015) have investigated the Danish case and came to a similar conclusion. However, systematic research whereby there was made use of cases of which was not known prior to the research that there was a high possibility of political influence, has never been done on governmental COI reports. It is highly important to investigate this relationship because COI reports produced by COI units, are the most important information sources that are used by asylum decision-makers (Gibb & Good, 2013, p. 294). And because the reports thus have such a high influence on the lives of asylum seekers, it is very important that these reports are objective. To find out if there is, as has been suggested by scholars, a relationship between COI reports and politics, I will try to answer the following research question in this thesis: ‘To what extent does domestic politics influence the production of Country of Origin Information (COI) reports?’

(7)

The findings of this thesis will not only have a social contribution but will also contribute to the existing literature. Over the years, academic scholars have found that political actors often try to influence the status determination process of their country, in accordance with their policy preferences. But most of this research has been done on the objectivity of asylum judges and policymakers (Keith et al., 2013; Kerns, 2000; Pettitt, 2009; Townhead, 2009; Huber, 2009; Gibb & Good, 2013; Thomas, 2008; Ramji-Nogales, Schoenholtz & Schrag, 2007; Aggarwal & Floridi, 2020; Tsangarides, 2010). However, this thesis looks into political influence on Country of Origin Information reports that are produced by countries. This source of information has gained importance in the last few years, and therefore scholars have only started researching it recently (Engelmann, 2015, pp. 110-11). There are multiple academic articles published on the relationship between COI and politics (Engelmann, 2015; Good, 2015; Gibb & Good, 2013; Van der Kist, Dijstelbloem & de Goede, 2019; Rosset & Liodden, 2015). But this research focuses most of the time on COI in general and not on COI reports specifically, while those reports have a special position within the status determination process. Besides, the research that has been done on COI reports that are produced by governmental COI units focused merely on the case of the Danish report on Eritrea (Van der Kist et al., 2019; Rosset & Liodden, 2015). The problem with those studies is that the scholars knew prior to the research that there was a high chance of political influence in that case. A structured research into the objectivity of COI reports in general has not previously been done, therefore the theoretical contribution of this thesis is that it will test if the existing theories (that there is a relationship between COI reports and policy preferences) are true for COI reports in general, or only in certain cases (for example the Denmark-Eritrea case).

Besides, this thesis also has a social contribution. If a relationship between COI reports and domestic politics is found, these results can be used to make the asylum and refugee status determination procedure fairer for asylum seekers. It is crucial to investigate this potential relationship because wrongful rejections of asylum claims could have significant consequences for the asylum seeker.

To answer my research question, I have conducted a comparative case study. COI reports from three reporting countries that differ in their level of restrictiveness of asylum and refugee policy were compared. On basis of the literature, it was expected that due to the relationship between politics and knowledge, these countries will produce different reports. I expected that the country with the most restrictive policy would produce more positive reports so that they have to grant less people asylum. Conversely, I expected that the country with the least restrictive

(8)

asylum and refugee policy would have more negative COI reports because on basis of that, they will accept more asylum seekers. I have collected data through desk research (COI reports) and through semi-structured interviews with individuals who are involved in the production and the use of COI reports.

Eventually the findings of the desk research suggested that there is a relationship between political preferences and COI reports. This would mean that the hypothesis (that there is a relationship) would be confirmed. The interviewees denied this relationship, which would mean that the hypothesis would be rejected. However, those results are questionable because the respondents worked at/with the Dutch COI unit and can therefore be biased. The main findings of this research are thus not conclusive about political influence in COI production, but definitely suggest that there is potentially such a relationship.

This thesis is structured as follows. First of all, I will explain what COI is, why it is important, what COI units are, and how these units produce reports. After this background information, I will give an overview of the existing literature on political influence in the status determination process to explain what is lacking in the existing literature and thus what the theoretical contribution of this thesis is. Thereafter, I will connect the literature review to my research in the theoretical framework, by explaining what relationship I expect to find between politics and COI production. After the presentation of the theoretical framework, the research methods will be discussed in chapter six. First, I will explain why I chose to carry out a comparative case study over other methods and what this will add to the existing literature. This will be followed by an explanation of the case selection method and why I chose certain cases. Thereafter, I will explain how my research will be conducted, how I will collect the data and eventually how I will operationalize the analysis of the reports to find proof for political influence. I discuss these subjects thoroughly because it is important for the validity and reliability of the research to use the correct methods and cases to investigate the expected relationship. These chapters will be followed by a presentation of the results, wherein I try to answer the research question. In that chapter, I will explain what I found during my desk research and how these findings can be explained, with the help of the conducted interviews. By dividing the result chapter in those sections, it will become clear if the existing literature is correct and if there is indeed a relationship between political influence and COI reports, or if this hypothesis will be rejected and the Danish case was an exception. Finally, in the last chapter of this thesis the main findings will be summarized, the limitation of the research will be explained, and I will make

(9)

recommendations for further research. After these chapters, the list of references and appendices follow.

3 Background information COI reports

Conceptualization Country of Origin Information

In this thesis, COI reports will be investigated. Therefore, ‘Country of Origin Information’ must be conceptualized. The official definition that is created by the Austrian Centre for Country of Origin & Asylum Research and Documentation (ACCORD), and used by, among others, the EU, is:

… information which is used in procedures that assess claims to refugee status or other forms of international protection. COI supports legal advisors and persons making decisions on international protection in their evaluation of the human rights and security situation, the political situation and the legal framework, cultural aspects and societal attitudes, the humanitarian and economic situation, events and incidents as well as the geography in claimants’ countries of origin (or, in the case of stateless people, countries of former habitual residence) or countries of transit. To qualify as COI, it is essential that the source of the information has no vested interest in the outcome of the individual claim for international protection. (ACCORD & Austrian Red Cross, 2013, p. 12)

Or, put more simply: ‘… information about the situation in refugees’ home countries which is used in procedures for determining international protection needs’ (ACCORD & Austrian Red Cross, 2013, p. 12). However, this definition can include many sources. Because it is hard to track all the different sources that can be used for asylum determinations, I will focus my research specifically on the reports that are produced by national COI units. Van der Kist et al. (2019) explain that ‘It is within COI units that information is … created, collected, assembled, transcribed, transported to, simplified and juxtaposed in a single location, … where everything that is relevant can be seen’ (p. 71). These units make use of different sources but produce their own reports which are in this thesis referred to as ‘COI reports’. It is very interesting to look specifically at these reports, because (almost all) COI units are part of a national government that has certain policy preferences and because of this special position within the government,

(10)

there could potentially be political influence on the production of these claimed-to-be objective COI reports.

Functions of COI reports

Country of Origin Information is crucial for the asylum process. On basis of the ‘principle of refoulement’, states are not allowed to send a person back to their country of origin if they fear persecution in that country. This principle was determined in the 1951 Geneva Convention and the 1967 Protocol Relating to the Status of Refugees, that ‘… recognized the right of persons to seek asylum from persecution in other countries’ (UNHCR, n.d., p. 2). Decision-makers determine who is at risk of persecution in the country of origin and therefore decide who has the right to asylum. For that procedure it is necessary to have information about an applicant's country of origin.

To support decision-makers with understanding the situation in a country of origin, COI reports are created. As explained before, these reports provide the decision-makers with information on a broad range of subjects. Because COI can be of great influence on the outcome of applications, it is highly important that COI is ‘… accurate, reliable, up-to-date, impartial and objective in order to guarantee a fair and credible examination of the applicant’s claim’ (International Centre for Migration Policy Development [ICMPD], 2006, p. 10). Besides, COI is also important for the people who apply for asylum because it helps them to make their claim, along with the people who (legally) support them.

COI units

Country of origin information is collected in all countries. However, the processes of collection and production are different between countries. As Engelmann (2015) explains:

‘On one end of the scale, it could be that an individual person employed by a national asylum authority has to decide about an asylum application and conducts in this context a cursory online research on the country of origin; on the other end of the scale could stand a fully-fledged research unit with people specialising in certain countries of origin, keeping elaborated and regularly updated files about the situation in these countries and providing these information to a different unit solely responsible for deciding on asylum applications.’ (Engelmann, 2015, p. 104)

(11)

COI units, which thus are the research units that produce the COI reports, differ from each other on many different levels. The ICMPD has compared the units from 10 different European countries. The main differences between the units were the ministry under which the unit operates, the number of employees in the COI unit, public access to the reports, the relationship with the policy-making department and the quantity of the production (ICMPD, 2006). Table 1 is focused on the three cases that will be researched in this thesis and provides an overview of this information, as well as some other relevant information such as the used languages and the name of the reports made by that particular unit.

The production processes

The production of COI can be a complicated process. One of the main problems in the production process of COI is that there are often many different sources that provide contrasting information, and the employees of the COI units have to decide which sources are reliable. To make the production of COI reports easier and more transparent, some organizations made guidelines for the production of COI reports, such as the ‘Common EU Guidelines for processing Country of Origin Information’ that is created by a number of EU member states, or the EASO Country of Origin Information report methodology, made by the European Asylum Support Office (EASO) (ACCORD & Austrian Red Cross, 2013, p. 39). One of the most commonly used guidelines is the COI training manual that is produced by ACCORD, with the support of the UNHCR, in which is explained how COI should be researched (ACCORD & Austrian Red Cross, 2013). They explain how to determine the reliability of a source. The most used external sources for COI are the media, academia, international and intergovernmental organizations, governmental organisations and non-governmental organisations and other civil society organizations (ACCORD & Austrian Red Cross, 2013, p. 87). To evaluate if a source is reliable, all the information of the different sources should be compared. Besides, the reliability of a source could be evaluated by answering questions about the information-gathering process. When information is not accessible through external sources, COI units can also produce the information used for COI reports themselves, by conducting a fact-finding mission where a team travels to the country that they are investigating (or the neighbouring country if visiting is impossible) to investigate the situation of that country.

(12)

Table 1: characteristics COI units2

Country Australia Netherlands UK

Name of the unit Country Information Report Section Country and Language Information Unit Country of Origin Information Service

Respective Ministry Department of Foreign Affairs and Trade (DFAT)

Ministry of Foreign Affairs Home Office

Division in COI production and policy making departments?

Yes Yes No

Number of staff in COI unit

n/a 24 (in 2012) 18 (in 2012)

Public access Yes, but only the reports that are currently used by the decision-makers

Yes, the most recent reports as well as the previous reports through the government archive

Yes, but only the reports that are currently used by the decision-makers

Name of reports Country Information Reports Algemene Ambtsberichten Country policy and information notes (previously known as country information and guidance reports)

Language(s) English Dutch and some collaborative reports in English

English

Production - 12 reports in 2019 - 32-36 country reports

- Reports on individual asylum cases

- Top 20 countries of origin twice a year

- 10 a year for countries with at least 120 applications

2 Sources:

Australia (Department of Foreign Affairs and Trade, n.d.)

The Netherlands (ICMPD, 2006; Engelmann, 2015; Ministerie van Algemene Zaken, n.d.; personal communication) The UK (ICMPD, 2006; Engelmann, 2015; UK Visas and Immigration, 2020)

(13)

During my research, there were three remarkable differences found between the structure of the reports of the three cases that are investigated in this thesis: Australia, the UK and the Netherlands. First of all, the Netherlands and Australia both produce one report per country that addresses all groups that are eligible for asylum, while the UK produces multiple reports by dividing the reports into different subjects. For example, the UK produced four reports on Somalia, not only on the overall security and humanitarian situation, but also more specific reports on clans, women fearing gender-based violence and fear of Al-Shabaab. For the case of Nigeria, this number is even higher. While the Netherlands and Australia produced one report, the UK produced nine.

A second noteworthy difference between the structure of the reports is the length. The reports from Australia are significantly shorter than the reports from the Netherlands and the UK. In the case of Nigeria, Australia has produced 34 pages of country information, while the Dutch report is 77 pages long. The UK’s Home Office stood out with their production of in total 389 pages of COI on Nigeria. The length of a report itself, however, does not necessarily proves anything about the presence of political influence, because a short report could support a countries’ policy preferences by giving more freedom to the asylum decision-makers to interpret the report and thus deny or accept more applications. Besides, a short report could also be less precise in who is or who is not in danger, and therefore include or exclude more people in the group that is allowed asylum.

One of the ways in which the differences in length can be explained is by the third found difference between the reports: referencing. The countries all used different methods to refer to the used sources. Australia, with the shortest reports, has in many cases not included a source on which their assessments are based. Also, no list with used sources is included in the reports. This saves a lot of text and shortens the reports, but also makes the reports less verifiable. On the other side of the spectrum, the UK has included citations of the used sources to show wherefrom they retrieved their information. In the first pages of their reports, they make their statements about the risks of different groups by referring to the citations in the second part of the report. The citation of the sources is the main reason why these reports tend to be so much longer than the other two. The referencing of the Netherlands can be positioned in between the former two. The country uses footnotes to indicate where it got its information from and has included a list of used sources in their reports.

(14)

3.4.1 COI cooperation

There has been an increased interest by countries (especially countries within the EU) to exchange their COI or to do collaborate fact-finding missions. There are two reasons for countries to support this cooperation. The first reason is that international cooperation moves the accountability away from the national government. Furthermore, it improves the consistency in asylum decision-making between countries and thus will lead to a more equal distribution of asylum seekers and refugees (Vink & Engelmann, 2012, pp. 546-547). There are many different organizations that collect and share COI (Vink & Engelmann, 2012, p. 547; Refugee Legal Aid Information, n.d.). The main global platforms for sharing COI or organizations that provide guidelines for COI production are Refworld, the UNHCR, Austrian Centre for Country of Origin and Asylum Research and Documentation (ACCORD), International Association of Refugee Law Judges (IARLJ) and the EU (Vink & Engelmann, 2012, p. 547; Refugee Legal Aid Information, n.d.; Gyulai, 2007). Nevertheless, as Vink and Engelmann (2012) explain: ‘an intergovernmental system for collecting, analysing and distributing COI remains underdeveloped’ (p. 546). Cooperation between countries experience obstacles, such as ‘… language and legal issues, different needs and focuses, quality standards, confidentiality issues’ (ICMPD, 2006, p. 10).

4 Literature review

COI literature in the broader context

Before I discuss the literature on country of origin information, I will touch upon the contribution of this thesis to some of the broader debates in International Relations.

First of all, there is an ongoing debate between realists and constructivists, called the ‘norms versus interests’ debate (Keith, Holmes, Miller, 2013, p. 263). Constructivists believe that states accept international norms because they are pressured by networks of actors and feel the legal obligation to accept norms, and therefore expect that the formal legal commitments to the norm of nonrefoulement have more influence on asylum outcomes than national interests (Risse, Ropp & Sikkink, 1999; Henkin, 1979; Chayes & Chayes, 1993; Keith et al, 2013). Realists, on the other hand, believe that states would only support a human rights norm (such as the norm of nonrefoulement), if it is in the state’s interest (Keohane, 1984; Waltz, 1997; Mearsheimer, 1994). If my hypothesis is confirmed, and COI is influenced by the state, this

(15)

could be a sign that realists are right because the state prefers to defend its national interests rather than to adhere to the international norm of nonrefoulement.

A debate that this research contributes more directly to, is the debate on the relationship between expert knowledge and politics. The definition of expert knowledge that will be used in this thesis is:

… the forms of codified knowledge that are either produced by specialists (as indicated by qualifications or institutional affiliation); or which involve specialist or technical methods, equipment or accumulated knowledge that is generally assumed to require skills and experience not possessed by professional administrators. (Boswell, 2017 in Littoz-Monnet, 2017a, p. 2)

While it appears that there is a boundary between expert knowledge and politics; experts produce knowledge, which is used by policymakers to inform and guide policy making, this separation is not so clear. Scholars have debated over the years about the relationship between expert knowledge and politics (Littoz-Monnet, 2017a, pp. 9-13). Some scholars believe that policy is shaped by experts through the knowledge they produce (Cannon, 2017), while others have researched if policymakers influence the production of knowledge (Littoz-Monnet, 2017b; Demortain, 2017; Berten, 2017). However, most of the scholars explain that this is an iterative process, and that experts and political actors both influence each other (Littoz-Monnet, 2017a; Freeman & Sturdy, 2017; Biersteker, 2017; Demortain, 2017).

In this thesis, COI reports will be researched. I believe that COI reports are a form of expert knowledge. Scholars such as Rosset and Liodden (2015) and Boswell (2009) share this understanding. COI is produced by specialists and provides decision-makers in asylum procedures with information that these decision-makers do not possess themselves. Therefore, the features of COI reports are in line with the aforementioned definition of expert knowledge by Littoz-Monnet (2017a). This research on the relationship between political and COI units contributes to the existing literature on expert knowledge by investigating whether the relationship is also existent when the knowledge is produced by governmental actors.

(16)

Objectivity in the status determination process

Objectivity is a very important aspect of the status determination process. The individual’s right to life is the most significant of all human rights and it is thus highly important that the basis of the asylum decision ‘fall for the most anxious scrutiny’ (Lord Bridge in Good, 2007, p. 129). The consequences of wrong assessment of an applicant’s claim could be significant.

If legal representatives, advisors or decision-makers are unable to properly access, understand and utilise country information, they cannot provide quality advice and representation to asylum seekers or make adequate decisions. Should the advice or representation given to asylum claimants be sub-standard or even erroneous, or if a wrong decision is made, the consequences for the individual could be significant. This might include enforced removal and ultimately return to a country where they could be at risk of persecution or death. (Tsangarides, 2010, p. 8).

In the whole status determination process, objectivity is important. Scholars have often researched the role of decision-makers in asylum procedures and criticized them for making contradictory decisions by phrasing the process as ‘refugee roulette’ (Tsangarides, 2010). The lack of objectivity of status determination decision-makers that use COI is discussed by many authors (Keith et al., 2013; Kerns, 2000; Pettitt, 2009; Townhead, 2009; Huber, 2009; Gibb & Good, 2013; Thomas, 2008; Ramji-Nogales, Schoenholtz & Schrag, 2007; Aggarwal & Floridi, 2020; Tsangarides, 2010). However, in this thesis, I am aiming to find out if political interests play a role in the production of COI reports. Some academics such as Good (2007) warned for the inaccuracy of COI reports earlier on (p. 213), but it was not until the last decade that scholars began to research the production of COI and its objectivity.

Divergence in Country of Origin Information

Multiple scholars have researched Country of Origin Information. COI can come from various sources. ‘They include generic information (for example, news bulletins and maps) as well as reports specifically produced and compiled for use in asylum proceedings…’ (Aggarwal & Floridi, 2020, p. 1). Some of the scholars researched COI in general (Engelmann, 2015; Good, 2015; Gibb & Good, 2013), while others focus more specifically on COI reports (Van der Kist, Dijstelbloem & de Goede, 2019; Rosset & Liodden, 2015). All of the articles criticize the use

(17)

of, or the production of, COI. The scholars that researched the production of COI reports came to the conclusion that COI reports differ between countries, while one would expect the reports to be similar because the reports all claim to be objective descriptions of the security situation in a country of origin (Good, 2007; Rosset & Liodden, 2015; Van der Kist et al., 2019). This thesis will contribute to the existing literature that tries to explain if and why COI reports differ between countries. More specifically, this research will focus on the possibility of political influence as an explanation for divergence in COI reports.

An explanation for the differences between COI reports could be that the experts that are involved in the production of COI are biased, and therefore it is very difficult to produce an objective report about a country of origin (Good, 2007, pp. 222-226). There are multiple forms of biases that COI producers can have. The first form is motivational biases, which are biases that occur when the producer has a preference for one side to win. In that case, the selection of certain information for COI can differ as a result of a producers’ preference for the asylum seeker to be granted asylum. Another form of bias that could play a role is cognitive bias. This is the unconscious preference towards information that favours the expectations of the COI producer instead of information that opposes the expectations. Lastly, ethical biases can occur when an employee of a COI production unit feels that it is their responsibility is to respect human rights which could lead to them being in favour of the asylum applicant’s side to win.

However, other scholars explained the differences between COI reports in a different way. While they do not rule out that biases of COI producers can be of influence, these scholars argue that the differences are the result of political influences (Aggarwal & Floridi, 2020; Tsangarides, 2010; Van der Kist et al., 2019; Rosset & Liodden, 2015).

Before we can understand how political actors could influence COI, it is important to explain why those actors would want to influence the reports in the first place. There are various reasons for political actors to get involved in the production of COI. To understand this, it is helpful to look at the functions of expert knowledge that Boswell (2008) described in her book. The first, and most obvious, function is the instrumental function of knowledge, which means that it is used to provide information. ‘Because they deal with highly complex and often technical issues, international decision-makers are highly dependent on science and technology for determining the risks and consequences associated with political action’ (Littoz-Monnet, 2017a, p. 7). And the more complex a problem is, the more important and influential expert knowledge is (Haas, 1992, p. 13). However, this is not only true for international

(18)

decision-makers. As Rosset and Liodden (2015) explain, this instrumental function is the intentioned purpose of Country of Origin Information reports; providing information for the asylum determination process (p. 27).

A second function of expert knowledge is that it can legitimize action (Littoz-Monnet, 2017a, p. 7). Boswell (2008) categorizes this as one of the symbolic functions of knowledge and explains that ‘by being seen to draw on expert knowledge, an organization can enhance its legitimacy and bolster its claim to resources or jurisdiction over particular policy areas’ (p. 472). Because of ‘… the perception that the bureaucrats possess reliable, relevant and detailed knowledge creates confidence that their proposals are well founded.’ (Littoz-Monnet, 2017a, p. 7). Governments also make use of this function. Especially in highly sensitive cases, such as asylum decisions, expert knowledge plays a significant role for political actors because ‘… decisions that reference COI are well-informed and rational, thus affirming the legitimacy of policy actors and institutions involved’ (Rosset & Liodden, 2015, p. 27) (Boswell, 2009).

Lastly, policy makers make use of expert knowledge because of its substantiating function (Boswell, 2009, pp. 72-81; Littoz-Monnet, 2017a, pp. 8-9). ‘Expert knowledge can lend authority to particular policy positions, helping to substantiate organizational preferences in cases of political contestation’ (Boswell, 2008, p. 472). Especially in highly political contested areas, policy makers try to find scientific authority to substantiate their policy preferences (Boswell, 2009, p. 62). This is different from the legitimizing function of knowledge. ‘In the case of legitimizing knowledge, an administrative agency is attempting to secure its legitimacy qua organization. It is keen to demonstrate its capacity to mobilize resources to produce and apply knowledge. In the case of substantiating knowledge, by contrast, the aim is to garner support for a preferred course of action.’ (Boswell, 2009, p. 73).

Rosset and Liodden (2015) explain very clearly how and why these functions of expert knowledge could also be applicable to COI:

Brekke (2004:45) suggests that politicians may use asylum policies as a means to create an impression of a restrictive country. Thus, in addition to the actual changes that a government puts into effect, policies have an equally important communicative and symbolic function. We suggest that COI can play a similar role in an asylum policy tool kit. First of all, COI has a direct impact on the number of applicants who are accepted as refugees. Secondly, when COI signals a restrictive turn, it may serve to deter asylum seekers’ decisions about a destination country. Eritrean asylum claims in Denmark dropped dramatically after the report was published—from 606 at their peak in August

(19)

20146 to just three in January 2015 (Danish Immigration Service 2014b, 2015)—which lends support to the hypothesis that the Danish authorities successfully communicated a restrictive turn to future applicants. (Rosset & Liodden, 2015, pp. 27-28)

Rosset and Liodden (2015) explain that more restrictive COI reports could also lead to a ‘race to the bottom’ (p. 28). If one country makes their COI reports more positive, and thus asylum policy more restrictive, this could encourage other countries to produce more restrictive reports as well. This could happen because if one country is less attractive for a group of asylum applicants, this leads to more applications in other countries. Those countries will then adjust their policies to avoid receiving the more asylum applications.

Now that we understand why policy makers would want to influence COI reports, we will continue with an explanation of how political actors could do this. Because of the aforementioned advantages that COI can have for policy makers, it is understandable that these actors will try to influence the production of COI reports to validate their policy preferences. While some academics only state that there is a possibility that COI is produced conform asylum policies (Tsangarides, 2010, p. 47), other scholars argue that COI units have become a crucial part of the policy shaping stage (Engelmann, 2015, p. 110). ‘… Country of origin information is needed far beyond the individual refugee status determination. Given the tendency to adopt asylum policies for groups of asylum seekers coming from certain countries of origin, this information increasingly becomes policy-relevant’ (Engelmann, 2015, p. 112). Furthermore, Van der Kist et al. ‘… argue that the relation between knowledge and decision-making is a process of translation in which facts and evidence are slowly consolidated to support decision-making processes’ (Van der Kist et al., 2019, p. 69). The information about countries of origin is often complex and arguably and can therefore be adapted in the production process to substantiate policy preferences. Van der Kist et al (2019) distinguish three phases in the production process to demonstrate how COI can be produced COI to substantiate a countries’ preferred policy: the phase of investigation, the concordance of information production and the phase of consolidation (p. 70). According to them, the production process takes place in the shadow of decision-making because COI units are not producing facts but produce information that enable asylum decisions (p. 80).

The first phase is the phase of investigation (Van der Kist et al., 2019, pp. 73-76). Employees of COI units gather information through desk research and fieldwork. In the case of desk research, producers have to rely on information from different online sources. COI

(20)

employees need to assess the reliability of a source and select between different sources. While most of the used sources have a certain kind of authority, the quality of the online sources often ‘… lacks proper channels of quality control’ (Van der Kist et al., 2019, p. 74). Fieldwork can be influenced in other ways. When a COI unit goes on a fact-finding mission, the validity of the information lies in the hands of the fieldworkers. This could be problematic for the gathering of information that is more open to interpretation. ‘Evidence on matters such as social impacts, human rights abuses, corruption, or violence are much more volatile, subjective, and open to cultural and linguistic (mis)interpretation and even fraud’ (Van der Kist et al., 2019, pp. 73-74). Besides, the results of a fact-finding mission can be influenced by practical problems such as the accessibility of a country. In both methods for information gathering, employees of COI units have the power to influence the outcomes of the findings. They can ‘… select, order, and present materials and achieve articulations that can be passed on to the next phase’ (Van der Kist et al., 2019, p. 75).

The second phase is the concordance of information production, which includes collecting, combining and comparing of information (Van der Kist et al., 2019, p. 78). The problem that arises in this phase is that the production and practices of COI differs within and between countries. There is little coordination in the production of COI and therefore the COI units can come to contradicting conclusions. There are various guidelines for the production of COI, but those are not enough to establish a homogenous production process for all reports in all countries (p. 77). To come to COI, producers gather information from different sources and combine those sources. However, this process is arbitrary because the ‘classifications and hierarchies of knowledge are arrived in a noncalculable way’ (Van der Kist et al., 2019, p. 78).

The last phase in the COI production is the phase of consolidation: ‘the authority of a COI report is drawn from its use in governmental deliberations, while decision-makers use the authority of country expertise to validate their policy decision … [In this phase] weak and wobbly country information becomes a strong and stable reservoir of authoritative governing knowledge’ (Van der Kist et al., 2019, p. 79). Information that is sometimes doubtful and incomplete, becomes qualified knowledge that is used for asylum policies (p. 79). This last phase shows how questionable information can become unquestionable facts.

(21)

Gap in the literature

The existing literature on COI has primarily focused on the objectivity of asylum decision-makers and the way in which they make use of COI. Literature on the objectivity of COI (reports) itself, has only established in the past decade. This thesis contributes to that literature. More specifically, it contributes to the literature on the relationship between COI production and policy preferences. It is essential to investigate the objectivity of COI reports that are produced by COI units because the importance of these reports in the asylum process has significantly increased over the years (Good, 2015, p. 123; Van der Kist et al., 2019, p. 69).

The existing articles on the objectivity of COI all found that there is a relationship between political actors and COI production (Aggarwal & Floridi, 2020; Tsangarides, 2010; Van der Kist et al., 2019; Rosset & Liodden, 2015; Engelmann, 2015). However, there are two shortcomings to the existing research on COI objectivity.

Firstly, some of the articles focused on COI in general, and thus not on the COI units that are part of the government and produce Country of Origin reports (Engelmann, 2015; Good, 2015; Gibb & Good, 2013). However, these units occupy a strategic position in the status determination process:

It is within COI units that information “is … created, collected, assembled, transcribed, transported to, simplified and juxtaposed in a single location, … where everything that is relevant can be seen” … COI units function as powerful nodal points where distant information, material, and specimens are gathered and combined, and where the truth of the situation in migrant-sending countries is inscribed. Consequently, COI unit practices enable “action at a distance” that allow the identification, sorting, and adjudication of asylum populations. (Van der Kist et al., 2019, p. 71).

It is important to look at this specific form of COI because the reports that are produced by COI units are the most important source of COI that asylum policy and decision-makers make use of: ‘HOPOs [Home Office Presenting Officers] draw their COI evidence almost entirely from the Country Reports produced by the Home Office’s in-house Country of Origin Information Service (COIS)’ (Gibb & Good, 2013, p. 294). Besides, the fact that COI units are (in almost all cases) part of the government that has certain policy preferences, makes its objectivity more

(22)

questionable than other sources that are produced by non-governmental sources. Because the COI units are part of the government, the outcomes of the reports can also be in their interest and can therefore be biased. This is also recognized by other scholars: ‘… this type of knowledge … is uneasily positioned in between social science and policy decision’ (Van der Kist et al., 2019, p. 69). Thus, because of its importance and its position, it is important to investigate this specific form of COI.

However, the literature that focuses on governmental COI reports is also lacking. Both Van der Kist et al. (2019) and Rosset and Liodden (2015) have merely researched one case to find out if and how the relationship between politics and COI units works. Not only have they merely looked at one reporting country (Denmark), but also at one country of origin (Eritrea). Both researches made use of the Danish report on Eritrea to substantiate their theories. In 2014, the Danish policy on Eritrean asylum seekers changed and Eritreans who were fleeing service in the national army and/or left Eritrea illegally ‘… would no longer constitute a ‘systematic risk of persecution upon return’’ (Rosset & Liodden, 2015, p. 26). This report was based on new country of origin information, that the Danish COI unit retrieved during two fact-finding missions. The report was criticized by different organizations (for example Amnesty International, Human Rights Watch and the UNHCR) because of methodological errors and disagreements that arose within the government (Rosset & Liodden, 2015, p. 26). Not only these organizations, but also the public suspected that political influence had taken place because ‘… it came at a time when there was considerable debate on the growing number of asylum applicants in Denmark’ (Van der Kist et al., 2019, p. 68). Both Rosset and Liodden (2015) and Van der Kist et al (2019) therefore researched if this was indeed case and why and how this had happened. But, researching a single case brings along limitations:

A “mere” case study … is often identified with loosely framed and nongeneralizable theories, biased case selection, informal and undisciplined research designs, weak empirical leverage (too many variables and too few cases), subjective conclusions, nonreplicability, and causal determinism. (Gerring, 2006, p. 6)

I believe that some of these limitations are also visible in the existing research on COI reports. While both articles acknowledge that this one case is not representing all COI reports, the scholars suggest that there might be a relationship between policy preferences and the production of COI reports:

(23)

The Eritrea case suggests that COI may function as a potential means of deterrence and that the boundaries between COI and policy goals blur easily, such that the production of knowledge becomes a site of political negotiation.

(Rosset & Liodden, 2015, pp. 27-28)

The Danish report on Eritrea cannot be considered representative of the COI work in all European countries, but it does illustrate the importance of paying close attention to the ways in which information is produced and used in a field that is markedly politicised (Rosset & Liodden, 2015, p. 29)

This suggestion is, however, based on one case of which the scholars knew prior to their research that there was a high chance of political influence. There is a possibility that this single case is an outlier, because of the biased case selection and the investigation of merely one case. Therefore, there must be done a research to find out if political influence of COI reports happens on a regular basis. A structural research to prove the existence of a relationship between politics and COI reports has never been done.

In this thesis, I am aiming to find out if political interests play a role in the production of COI in general, or whether the case of Eritrea was an exception. I will conduct a comparative case study whereby I will compare three different reporting countries and two different countries of origin to find out if this theory holds when cases are chosen of which it is not know in advance that there are differences between the reports. By using this case selection method, I will try to find out if the Eritrean case was an outlier or if there is indeed a relationship between COI reports and politics in general.

5 Theoretical framework

The relationship between policy and COI reports

In this section I will explain how the expected relationship between asylum and refugee policy and COI reports works and on which literature this expectation is based. In the literature review, I explained that COI reports can substantiate asylum policy in two ways. Firstly, COI reports can directly influence the number of asylum applications that are accepted. Besides, when COI reports support a more restrictive policy, this could lead to a lower number of overall

(24)

applications because in that case asylum seekers prefer a country where they are more likely to be granted asylum (Rosset & Liodden, 2015, pp. 27-28). This means that countries with a restrictive asylum and refugee policy will favour COI reports that will lead to the lowest number of acceptances and will discourage individuals to seek for asylum in their country.

In this thesis I will categorize the reports from ‘most positive’ to ‘most negative’. I expect that the most restrictive country will have the most positive report, and that the least restrictive country will have the most negative report. A ‘most positive’ report means that the reporting country is the most positive about a country of origin’s security situation. When this is the case, the group of individuals who can claim that they are at high risk of persecution in their country of origin, and thus fall under the principle of nonrefoulement, is lower. The reporting country is in that case, according to their information, not obliged to grant these individuals asylum. Rosset and Liodden (2015) explain how this can lead to a lower number of acceptances and a lower number of asylum applications.

A similar reasoning can also be used for the ‘most negative’ reports. I expect that countries that have a less restrictive policy are willing to accept more asylum applications, and therefore will produce reports that presume a more negative security situation. In that case, more individuals can claim that they are at risk at their country of origin and will thus be granted asylum. I understand that the terminology of ‘positive’ and ‘negative’ can be confusing, because a positive report can have negative consequences for the asylum applicant. However, this terminology was the best option I could find to describe the differences between the reports and therefore I’ve chosen to stick with it.

The expected relationship is that more restrictive countries will prefer more positive reports, and the other way around. Therefore, the existence of political influence in the production of COI reports can be researched by investigating if there is a relationship between the restrictiveness of asylum policies of reporting countries and the positivity/negativity of their COI reports. Proof for my hypothesis, which is that there is a relationship between political preferences and COI reports, would thus be if the most restrictive country has assessed that less categories of people are in danger of persecution than the least restrictive country has assessed. That would mean that a lower number of people can make a claim to asylum in the most restrictive country (which is in accordance with their policy preferences) than in the least restrictive country. A full explanation of how this analysis of measuring positivity or negativity of a report can be found in the section on the operationalization in the following chapter.

(25)

6 Research methods

Methodological approach

The ideal method to find an answer to my research question would have been to look at all the reports of all the countries that write COI reports and process trace the production of these reports, from the information gathering phase to the publishing of the final report. In that way I would find out where and how political actors have contact with the experts and see if that contact has led to changes in COI production. However, because COI information is conducted from many sources, this process is very hard to trace. Besides, looking at all the countries and all the reports would lead to an enormous amount of data, and thus would leave me with less time for an in-depth analysis to find the underlying reasoning for the differences in COI reports. Furthermore, a longitudinal research would have also been an option. I would then investigate if changes in political preferences over time, would lead to changes in the reports about the countries of origin. However, I have chosen not to do this because not only the domestic politics change over time, the security situation in a country of origin can change over time too. It would therefore be very hard to find out if changes in COI reports are the result of political shifts in reporting countries or in countries of origin. Besides, COI reports are written over a longer period of time and the production process could therefore overlap different governments in power, which would make it difficult to prove that the differences in COI reports are the result of a government’s political preference.

Thirdly, I chose not to study the decision-making process through ethnography because the main question that I would like to answer with this thesis if the production of COI is influenced by politics, and not how. Van der Kist, Dijstelbloem and De Goede (2019) and Rosset and Liodden (2015) looked into the production process of COI in Denmark to find where political influence took place. They chose Denmark because they (and the Danish society) suspected that the report on Eritrea was influenced by politics and wanted to find out how this happened. However, in this thesis I am aiming to find out if COI reports are influenced by political preferences in general, and I thus want to provide a more generalizable answer by looking at multiple cases and doing a systematic research. Another reason why I am not looking into the decision-making and information-gathering processes of COI is because it is a very complex process that is hard to trace because the production of COI makes use of many sources and it is hard to get access to insights about the production-process of the reports in COI units.

(26)

Another possible research method would have been a comparative case study of two cases (UK and Australia). While this method was my first choice, I found that there is limited availability of COI reports and/or drafts and therefore there would not be enough material to compare. Besides, more cases will lead to a more generalizable outcome and will increase the overall validity because by testing multiple reports, there is a smaller chance the cases are outliers and the found differences are, for example, the result of other variables like geography.

Instead of the abovementioned research methods, I will conduct a comparative case study of three reporting countries on two countries of origin, to find out if domestic politics influence the production of COI reports. By using this type of research method along with a structural method of analysing the reports, I will try to prove if, and to what extent, a relationship exists between political preferences and the reports. Previous research on this relationship has merely focused on one case, and therefore proof of general political influence is lacking. Comparing cases that are selected on the basis of their policy preferences, will lead to results that say more about the COI reports in general, than the existent single-case research can.

I only focus on the production of the reports within the units because, as Van der Kist et al. (2019) found in their research, the information-gathering phase consists information from many different sources and is therefore hard to investigate. I will compare three cases out of which two differ the most on the restrictiveness of their asylum and refugee policy, because this is where I expect to see the most differences in COI reports. Australia and the UK are chosen, as will be explained in the following section, because they have the least and most restrictive asylum and refugee policies. The third case, the Netherlands, can be placed in between Australia and the UK on its level of restrictiveness. This third case is used to control if Australia and the UK are not outliers, and thus makes the research more generalizable. The same reasoning applies to the choice to investigate of two countries of origin. If I would only research one case, it could be possible that this one case is an outlier. By researching three cases, I can research the reports thoroughly, while also being able to say something about the reports more generally. Based on the literature, my hypothesis is that domestic politics does influence the production of COI, and therefore I expect that the report of Australia will be the most positive, the reports of the United Kingdom will be the most negative and the Netherlands will be in between those two.

(27)

Case selection

In this section, the selection of cases will be discussed. After a description of the case selection method, the selection procedure will be explained. The case-selection of this thesis is divided into two parts. First, the choice of which reporting countries will be investigated in this thesis will be explained; thus, the countries that write the COI reports. Thereafter, the countries of origin about which the reports are written will be selected. These choices will be discussed consecutively.

6.2.1 Case selection method

The case selection method that will be used in this thesis is the most-similar cases design. According to Gerring (2006), there are two different uses for this design: hypothesis generating and hypothesis testing. For hypothesis generating studies, the researcher chooses cases that are on every aspect similar but differ on the dependent variable. In that way, the researcher can find out what independent variable is deviating and thus what variable is responsible for the different outcomes (Gerring, 2006, p. 131). According to Gerring, the most-similar cases design can also be used for hypothesis testing research. When a theory expects that a difference in the independent variable leads to a difference in the dependent variable, this case selection method can be used to test the theory (Gerring, 2006, p. 131). The selected cases would then, ideally, be similar in every respect, but different in one of the independent variables to find out if this leads to differences in the dependent variable. This thesis makes use of this second purpose, hypothesis testing. On basis of the existing theories, I expect that domestic politics does influence the production of COI. To test this hypothesis, I will choose three cases that are similar in most respects but differ on the independent variable ‘domestic politics’.

Gerring (2006) discusses that there are multiple limitations to this case-selection method. Firstly, he states that ideally the selected cases would be completely similar to each other. However, when investigating politics, this is often impossible because there could be a countless number of independent variables that influence the dependent variable (p. 135). Therefore, he explains that ‘In situations where exact matching is infeasible, researchers may instead employ approximate matching, where cases from the control group that are close enough to matching cases from the treatment group are accepted as matches’ (Gerring, 2006,

(28)

p. 135). While ‘… the definition of “close enough” is inevitably arbitrary …’ (Gerring, 2006, p. 135), I will try to choose cases that are similar on as many relevant aspects as possible.

A second limitation of this case-selection method is that the research can be biased if the cases are two outliers, and thus non-representative. Gerring explains that this problem can be overcome by ‘… assuring a choice of cases that are not extreme outliers, as judged by their residuals in the full model’ (Gerring, 2006, p. 139). I will argue in the following part about the selection of reporting countries that my cases are not outliers on their restrictiveness on asylum and refugee policy spectrum and will use a third case as a control variable that is in the middle of the two cases.

6.2.2 Selection of reporting countries

In this section I will explain which cases I have chosen for my research and why. At the end of my research, I want to be able to say something not only about my chosen cases, but also more generally about my ‘universe of cases’. The ‘universe of cases’ in this thesis is the production units of COI reports.

To find out if domestic politics influence the production of COI reports, I have compared three countries that differ on their asylum and refugee policy. I have chosen this way of measuring political influence because I expected that a more restrictive asylum/refugee policy would result in a more positive COI report, so the state can deny more asylum applications, and a more unrestrictive asylum/refugee policy would result in a more negative COI report, so the state can accept more asylum applications.

To find out what countries are at the ends of the asylum and refugee policy spectrum, I have used data from the Immigration Policies in Comparison (IMPIC) database (Helbling, Bjerre, Römer & Zobel, 2017). Ideally, I would have used the International Migration Policy And Law Analysis (IMPALA) database, which is a ‘national, institutional, cross-disciplinary project on comparative immigration policy’ and is more detailed and focused than the IMPIC database (International Migration Policy And Law Analysis, n.d.). Unfortunately, this database is hard to access and therefore I made use of IMPIC. The IMPIC database measures, among other things, the restrictiveness or non-restrictiveness of asylum and refugee policy. The researchers collected their data through surveys where country experts had to answer questions about five aspects of their countries’ legally binding asylum and refugee regulations: eligibility and accessibility to obtain asylum and a refugee status, conditions for

(29)

applying for asylum, the security of the status, the rights associated with asylum and refugee policies and the control of the migration policy (Helbling, Bjerre, Römer & Zobel, 2016, p. 9). On basis of the results, IMPIC came up with an asylum/refugee policy index, where countries receive an index number between 0 and 1, where 0 is the least restrictive index number, and 1 the most restrictive (Helbling et al., 2017). This provided me with a list of countries, ranked from least restrictive to most restrictive asylum and refugee policy (Table 2). The highest index number is 0.5 (Israel) and the lowest index number is 0.1 (Australia). The list shows that the index numbers are more or less equally divided, and the lowest and highest numbers are thus not outliers. This solves the problem of unrepresentativeness that was discussed in the section on the case selection method. Important to note is that asylum and refugee policy is different from border control. That explains why Australia has the lowest index number, which might appear odd. The country is generally seen as an anti-refugee country, because they have strict border control to prevent refugees from illegally entering their country. However, if a refugee asks for asylum in a legal way, Australia has one of the least restrictive asylum and refugee policies and therefore has the lowest index number on this topic in the IMPIC database.

Table 2: Country ranking asylum and refugee policy index3

Country IMPIC index (until 2010)

Australia 0.1 Portugal 0.1 Canada 0.2 New Zealand 0.2 Spain 0.2 Switzerland 0.2 Austria 0.2 Sweden 0.2 Korea (South) 0.2 Belgium 0.3 Ireland 0.3 Netherlands 0.3 France 0.3 Luxembourg 0.3 Italy 0.3 Denmark 0.3 Norway 0.3

3 Source: Helbling, M., Bjerre, L., Römer, F., & Zobel, M. (2017). Measuring Immigration Policies: The IMPIC

(30)

Finland 0.3 Poland 0.3 Czechia 0.3 Slovak Republic 0.3 Hungary 0.3 Estonia 0.3 Turkey 0.3 Japan 0.3 Iceland 0.3 United Kingdom 0.4 United States 0.4 Mexico 0.4 Chile 0.4 Germany 0.4 Greece 0.4 Israel 0.5

Based on this list (Table 2), my ideal cases would be Australia and Israel. However, there were some limitations to my case selection. Firstly, due to the language barrier, I could only choose cases that write their COI reports in English or Dutch. Translation was not an option because I am doing an in-depth case study, and therefore it is necessary that word choices and details do not get lost in translation. Furthermore, the reports needed to be accessible, for the obvious reason that I could otherwise not research them. Another limitation is that the country that I chose needed to write their own reports. Some countries make use of reports that are written by COI units from other countries and in that case, the reports are not the result of the (possible) relationship between politics and COI production (Engelmann, 2015, p. 106). Lastly, the countries needed to make individual determinations on asylum and refugee statuses because those are the cases in which the COI reports are of great influence.

These limitations left me with the following set of cases to choose from: Australia, Belgium, Canada, Ireland, the Netherlands, New Zealand, the United Kingdom (UK) and the United States (US). Two of the cases needed to be the two countries that are on the ends of the spectrum of the asylum and refugee policy index. On the one side of the spectrum is Australia, that has, with an index number of 0.1, the least restrictive asylum and refugee policy, and therefore is my first case. On the other side of the spectrum are the UK and the US, that both have an index number of 0.4 and therefore have a much more restrictive policy than Australia.

COI reports: objective truth or political tool? The influence of domestic politics on the production of Country of Origin Information reports