Psychological assumptions underlying credibility assessments in Finnish asylum determinations

Abstract

Credibility assessments are an important but complex part of asylum procedures. The current study investigated psychological assumptions underlying credibility assessments in Finnish first-instance asylum procedures and how these assumptions fit with widely accepted psychological science. Following previous research, we categorized assumptions in 56 real-life asylum cases from the Finnish Immigration Service. We found that asylum officials held assumptions about how truthful applicants present their claims, the plausibility of individuals’ behavior in their home countries, and applicants’ knowledge about asylum procedures. The assumptions were only partially in line with psychological science on memory, trauma, intercultural communication, and decision-making. To improve decision-making, training programs for asylum officials should include relevant findings from psychological science. To increase the transparency and combat bias, the written determination letters should also include explicit information about the decision-makers’ reasoning processes.

Keywords:

• Asylum seeker
• asylum procedure
• credibility assessment
• credibility indicator
• decision-making

Introduction

Deciding whether a person is in need of international protection or not has been described as one of the most complex and difficult forms of decision-making in the modern state (Thomas, 2006). Despite the clear theoretical definition and legal criteria for granting asylum status outlined in the 1951 United Nations convention and the EU Qualification Directive, the process of determining eligibility is in practice complex (Qualification Directive, The European Parliament, 2011; United Nations, 1951). Contrary to claimants in other legal cases, asylum seekers rarely possess documents that would substantiate their claims (Kagan, 2003). Because of this, asylum seekers are only required to verify their asylum claims to a reasonable degree of likelihood, instead of beyond a reasonable doubt (Thomas, 2006). The principle of the benefit of the doubt also states that, if certain conditions are met, decision-makers should accept an applicant’s statements even without corroborating evidence (for more information, see, e.g. Qualification Directive, The European Parliament, 2011). According to the EU Qualification Directive, The European Parliament (2011), the asylum seeker has to provide the authorities with all relevant information, whereas the official must assess and ascertain the information in cooperation with the applicant. In the credibility assessment, the believability of each element should be assessed impartially and objectively. Because only the elements that are perceived as believable are forwarded to the substantiality assessment (i.e. the process of deciding whether the conditions for granting asylum are met), the credibility assessment is of key importance in the asylum decision-making process (Kagan, 2003; Qualification Directive, The European Parliament, 2011). Several jurisdictions, including the United Nations High Commissioner for Refugees (UNHCR) and the International Association of Refugee Law Judges (IARLJ), have developed detailed guidelines for conducting fair and reliable asylum procedures (IARLJ, 2015; UNHCR, 2013). Despite the detailed guidance on how to conduct credibility assessments, decision-makers continue to base their decisions on unreliable markers of credibility such as demeanor and delay (Baillot et al., 2014; Dowd et al., 2018; Granhag et al., 2005; Herlihy et al., 2010; Kagan, 2003). The current study replicated and extended previous studies by identifying assumptions that underlie credibility assessments and by comparing these assumptions against widely accepted psychological science.

Current best-practice guidelines for evaluating asylum claims

The latest guideline from the UNHCR includes a thorough analysis about how credibility assessments are conducted within the European Union (EU). It also provides detailed information about credibility indicators and how they should be implemented in practice (UNHCR, 2013). Officials are encouraged to use five credibility indicators when judging the believability of statements: sufficiency of detail and specificity, internal consistency, consistency with other witnesses, consistency with information from external sources, and plausibility. Although helpful, these indicators are based on the assumption that applicants are able to recall and present an account of their past experiences in a detailed, consistent and plausible manner. This is not fully supported by empirical research (Dowd et al., 2018; Granhag et al., 2017; Herlihy & Turner, 2009). Several authors have criticized the criteria of “plausibility” for being based on fundamentally subjective and culture-specific commonsense judgements (Herlihy & Turner, 2009; Maegherman et al., 2018; Sweeney, 2007). Moreover, the criteria of sufficiency of detail and consistency have been criticized for being subjective when applied, despite conceptually seeming objective, as well as being too extensively relied upon (Maegherman et al., 2018; Sweeney, 2007). The analysis in the UNHCR handbook (2013) agrees that asylum officials sometimes make decisions on premises that are in conflict with current psychological knowledge about for example, memory, cultural differences, and decision-making. Decision-makers should, therefore, be informed about empirical psychological evidence regarding individual and contextual circumstances that might affect the asylum procedure, and carefully take these into consideration in their decision-making (UNHCR, 2013).

Psychological literature relevant to asylum procedures

Empirical research in psychology provides knowledge about several aspects relevant to asylum procedures, such as memory functions, the impact of trauma and culture on memory, intercultural communication, and decision-making. As in-depth literature reviews on most of these topics exist (e.g. Cameron, 2010; Granhag et al., 2017; Herlihy et al., 2012; Herlihy & Turner, 2009), we provide only a brief overview. In this overview, we focus primarily on psychological research.

The variation and limits of memory

Autobiographical memory (i.e. the memories of personally experienced events in an individual’s own life), as memory in general, is not an objective, exact recollection of events, but rather a constructive process prone to distortion and errors (Conway & Loveday, 2015; Hyman & Loftus, 1998). The details attended to and encoded into memory depend on an individual’s previous knowledge and experiences (Cohen, 2001). Memory for repeated events are blended together, making it difficult to remember specific instances (Herlihy et al., 2012). Moreover, the central gist of an event is often remembered, whereas peripheral information (e.g. temporal information such as dates, frequencies, durations and sequences, the appearance of common objects, names and verbatim wording of verbal exchanges) may be lost (Cameron, 2010). Furthermore, each retelling of a memory differs from another. People tend to recall more information with repeated recalls (i.e. hypermnesia; Cameron, 2010; Cohen, 2001). Social demands and suggestive questioning, on the other hand, might lead a person to develop and report distorted or entirely false memories (Hyman & Loftus, 1998). In sum, inconsistencies, vagueness, and gaps in recall occur naturally in both true and false recollections.

Trauma and memory

Research on how trauma and trauma-related disorders such as Post-Traumatic Stress Disorder (PTSD) affect memory is extensive but not without controversies (Engelhard et al., 2019). While evidence show that traumatic memories have specific characteristics, such as a tendency to be dominated by sensorial/perceptual and emotional details, the belief that traumatic memories are processed by special mechanisms or fragmented and incoherent by nature is not supported by evidence (Crespo & Fernandez-Lansac, 2016; Engelhard et al., 2019). While some studies have found that traumatic memories are more vulnerable to distortions compared to neutral memories (Southwick et al., 1997), others have found that traumatic memories are recalled more accurately than neutral memories (e.g. Porter & Peace, 2007). Most researchers currently agree that traumatic memories are prone to similar distortions and errors as neutral memories (Strange & Takarangi, 2015). A recent meta-analysis indicates that high stress may impair memory when it occurs before and during encoding, or at retrieval (Shields et al., 2017). Similarly to neutral memories, the central gist of traumatic experiences is usually remembered, but peripheral details may be forgotten (Moore & Zoellner, 2007). Additionally, persons exposed to trauma have a tendency to report less specific autobiographical memories, leading to overly general memories (Barry et al., 2018). Individuals might also, consciously or unconsciously, avoid traumatic memories to avoid unwanted emotional states (Herlihy et al., 2012). Whereas much of the research has focused on PTSD, depression and anxiety may also disrupt memory functions, for example, by impairing the efficacy of attentional control or leading to an overly negative recall of memories (Chaudhary, 2010; Derakshan & Eysenck, 2009). Because asylum seekers have a higher prevalence of mental health issues than the normal population (Fazel et al., 2005), these aspects are important to consider.

Memory and culture

Depending on their cultural background, people store and recall autobiographical memories differently (Gutchess & Indeck, 2009; Wang, 2016). Differences have been found in whether memories are self-focused or focused on social interactions, how detailed and specific memories are, the valence (positive vs. negative emotional characteristics) of memories, the emergence of early autobiographical memories, as well as the accessibility and functional use of memories (Wang, 2016). As could be expected, persons from more individualistic cultures tend to provide specific, self-focused, lengthy memories focused on individual experiences, whereas persons from collectivistic cultures tend to provide more general descriptions, focusing on collective activities, social interaction, and the role of others (Herlihy et al., 2012; Nelson, 2003). Asylum officials should be aware of these variations when assessing the narratives of applicants from different cultures.

Intercultural communication

Successful intercultural communication is characterized by participants actively trying to understand each other, as well as mutual trust (Gyulai, 2013). Challenges to successful intercultural communication are often due to differences in communication style and in the use of words, notions, and concepts (Granhag et al., 2017). Research on communication styles has distinguished high-context (e.g. Asian) cultures and low-context (e.g. Western) cultures, depending on how explicitly messages are exchanged and how much the context is expected to contribute (Hall, 1976). Recent studies have shown that persons from high-context (vs. low-context) cultures tend to report less details in response to open, free recall questions in interviews (see, e.g. Vrij et al., 2021). Furthermore, asylum seekers may understand instructions to “tell the truth” or “tell everything” differently from the official, and applicants might have difficulties knowing which details are relevant to their cases (Granhag et al., 2017). The way in which individuals present their narratives differ in both verbal directness (direct or indirect) and emotional expressiveness (restrained or articulate; Granhag et al., 2017). Finally, applicants from cultures with large power distances (i.e. hierarchical standards defining how authorities are perceived and approached) might find it inappropriate, or even shameful, to present objections or additions to asylum officials (Granhag et al., 2017; Herlihy & Turner, 2009).

Factors affecting decision-making

Much of the psychological research on human decision-making is relevant to the asylum context, although only few studies have specifically focused on asylum decisions (Herlihy & Turner, 2009). Human decision-making is prone to subjectivity on different levels. The cognitive system relies on shortcuts, or heuristics, and this makes human reasoning vulnerable to error and bias (e.g. Dror & Charlton, 2006). Tversky and Kahneman (1974) described two important heuristics: the availability and representativeness heuristics. With these heuristics, humans reduce complex tasks to simpler judgements based on what comes easily to mind and on similarity, rather than on probability. Another important heuristic, the confirmation bias, describes our tendency to seek information that confirms our belief and disregard evidence against it. The primacy effect and belief perseverance again represent our tendency to give undue weight to our first impressions and adhere to initial beliefs even in light of new evidence against them (Burke, 2006; Kassin et al., 2013). This tendency has also recently been found in asylum decision-making (Maegherman et al., 2018). Importantly, bias affects everyone and is not only due to incompetence (Dror, 2020). Experts may even be more susceptible to bias, especially in fields where the ground truth is impossible to know (Dror, 2020; Dror & Charlton, 2006). In such fields, the accuracy of decisions cannot be evaluated, and therefore, decision-makers can seldom learn from mistakes or outcome feedback (Dror, 2020; Pompedda et al., 2017). Considering the uncertain evidentiary context of refugee status determination, asylum adjudicators may be especially prone to using heuristics in their decisions.

Likewise, individual and contextual circumstances may affect decision-making (Dror, 2020). Studies have found that the mood, mindset, and ideology of decision-makers can affect judgements (Danziger et al., 2011; Giner-Sorolla et al., 2002; Herlihy & Turner, 2009). Furthermore, generalizations and stereotypes are helpful when navigating the complex social world, however, they are based on our limited personal or second-hand experiences and might distort our decision-making (Herlihy & Turner, 2009). In the asylum context, a large-scale study of asylum cases in the United States revealed large discrepancies in grant rates between asylum adjudicators, depending on individual characteristics of the immigration judge, such as gender and work experience (Schoenholtz et al., 2007).

Therefore, to combat bias, decision-makers must acknowledge the fundamental tendency of the human mind to be biased and to systematically check for potential biases in their own decision-making.

In conclusion, to properly employ the recommended credibility indicators and assess the credibility of asylum applicants and their narratives, it is vital that asylum officials are informed about the aforementioned topics in psychological science.

Previous research on assumptions underlying credibility assessments

Previous research has addressed various problematic aspects of credibility assessments in asylum cases (see, e.g. Cameron, 2010; Cohen, 2001; Granhag et al., 2005; Kagan, 2003; Maegherman et al., 2018). Nonetheless, only two studies have explicitly analyzed the underlying assumptions held by decision-makers in written determination letters (Dowd et al., 2018; Herlihy et al., 2010). Herlihy et al. (2010) qualitatively analyzed British immigration judges’ assumptions in decision-making in a sample of 30 written asylum determinations. Through an inductive thematic analysis, the authors identified three themes: There: how others behave, Here: the asylum system, and A truthful account, each with sub-themes. The first theme consisted of assumptions about how individuals, families and authorities behave in their country of origin. The second theme described assumptions about the appellants’ knowledge of and ability to navigate through the asylum system, and other professional actors’ clinical judgements. The third theme consisted of assumptions about the characteristics of truthful testimonies, such as the appellant’s demeanor. The authors also briefly evaluated these assumptions against empirical evidence and highlighted the need for more cross-disciplinary research.

In 2018, Dowd and colleagues analyzed decision-makers’ assumptions in 50 asylum cases processed by the Australian Refugee Review Tribunal (Dowd et al., 2018). The aim was to outline the extent to which Tribunal members acknowledge credibility guidelines and psychological science when applicants’ claims cannot be substantiated by corroborating evidence. The authors used the same themes as Herlihy et al. (2010), but partially adapted them to better fit their sample (Table 1). The first theme, Now: The applicant as a truthful witness, included three sub-themes of assumptions about the manner in which the applicant’ presented their claims, the consistency of the claims throughout the process, and the specificity of the claim. The second theme, There: Plausibility of account, included four sub-themes of assumptions regarding the plausibility of the behavior of both applicants, family members and State or non-State actors, and the characteristics of durable threats. The third theme, Here: The asylum system, included two sub-themes of assumptions about the applicants’ knowledge of and behavior in the asylum system in the receiving country, and the use of the benefit of the doubt-principle.

Table 1 Themes and sub-themes of assumptions made by asylum officials (Dowd et al., 2018).

Now: The applicant as a truthful witness	There: Plausibility of account	Here: The asylum system
Demeanor	Assumptions about the conduct of State and non-State actors	Assumptions about delay and disclosure
Inconsistency	Assumptions about threats being only temporary	The benefit of the doubt
Detail	Assumptions about fear, fight, or flight
	Assumptions about family

Dowd et al. (2018) found that Tribunal members approach the credibility assessments inconsistently, and that many assumptions contradict widely accepted psychological knowledge about memory and behavior—especially the assumptions about demeanor, consistency, and detail. The authors concluded that such contradictions substantially impact the validity of asylum determinations, and recommended Tribunal members to strive for predictability and consistency in their decision-making.

The current study

More research that systematically examines decision-making in real-life asylum decisions is needed to improve credibility assessment guidelines and address current challenges in using credibility indicators. Our main aim with the current study was to identify underlying assumptions and expectations in credibility assessments held by Finnish asylum officials. We also investigated whether the identified assumptions correspond with widely accepted psychological science on memory and behavior. Importantly, in contrast to previous studies on appealed asylum cases, the current study extended the field of study by using a randomized sample of cases processed in 2017-2018 from the Finnish first-instance asylum authority. Following the previous research by Herlihy et al. (2010) and Dowd et al. (2018), we first established coding principles for identifying assumptions and then assigned assumptions to the themes. As most assumptions fit the themes described in previous research, we decided to use the same themes and sub-themes as Dowd and colleagues in our coding. The small number of assumptions that did not fit those themes were examined thoroughly and given new labels.

Methods

Ethical permission

The study received permission by the Research Ethics Board for Psychology and Speech and Language Pathology of Åbo Akademi University.

Case selection

We included 56 cases from a randomly selected sample of 200 official asylum cases obtained from the Finnish Immigration Service in 2019. The case files were classified as publicly available and provided to the researchers after anonymization. The cases were processed between 2017 and 2018 in the four active units of the Immigration Service in Finland at the time. Each asylum casefile included transcripts of all available asylum hearings and the written decision. The applicants were both adults and minors, adults applying individually, or with their families.

To be included, a case had to contain: 1) a concluding determination (46 expired or “not investigated” cases were excluded) and 2) explicit statements about the applicant’s credibility (98 cases were excluded). The final sample consisted mostly of rejected cases (n = 47). The percentage of rejections did not reflect the general ratio of rejected to granted cases in Finland but was due to the fact that Finnish asylum officials are not obligated to explain their reasoning in the written determination letters of granted cases and were thus insufficiently informative for the current study. Of the nine granted cases included in the sample, one received asylum status and eight subsidiary protection status or a residence permit on other grounds.

Procedure

First, we familiarized ourselves with how Finnish asylum officials formulate their conclusions and structure their decisions. We focused on the section in the determination letters called “facts”, as there was no separate section explicitly called “credibility assessment”. In the “facts” section, officials mainly described credibility issues but also included conclusions about the substantiality assessment (i.e. whether the presented facts are sufficient grounds to be granted asylum). Second, we established the coding principles for identifying assumptions in the determination letters. Third, we read through all determination letters and entered all assumptions to a data file. Fourth, we assigned preliminary labels to all assumptions in the data file. At this point we noticed that almost all assumptions fit the themes and sub-themes described in the previous research by Dowd and colleagues. Therefore, we decided to use the same themes and sub-themes as Dowd and colleagues in our coding of theme. The few assumptions that did not fit any of the previously defined sub-themes were later categorized into new sub-themes.

Definitions and coding principles

An assumption was defined as a statement, or statements, about a credibility indicator or any other fact affecting the credibility assessment of the applicant’s claim, in which the asylum officials explained their conclusions (e.g. “Your story about the kidnapping lacks details and specificity. Therefore, the Finnish Immigration Service does not believe that it reflects a true event.”). To minimize coding bias, an assumption had to include both a statement about a credibility aspect or fact and a statement about the interpretation or conclusion of how this affected the decision. Consequently, we did not code vague statements not explicitly connected to facts or conclusions as assumptions (e.g. “You have reported things inconsistently.”). Nevertheless, it is important to acknowledge the subjective nature of coding assumptions when interpreting the findings.

We used the same set of themes and sub-themes described by Dowd et al. (2018): 1. Now: 1.1 Demeanor, 1.2 Inconsistency, 1.3 Detail; 2. There: 2.1 Assumptions about the conduct of State and non-State actors, 2.2 Assumptions about threats being only temporary, 2.3 Assumptions about fear, fight, or flight, 2.4 Assumptions about family; and 3. Here: 3.1 Assumptions about delay and disclosure, and 3.2 The benefit of the doubt (see Table 1). We later revised the names of the sub-themes to better fit the current sample. We coded assumptions that fit several sub-themes into all relevant sub-themes, for example, the statement “Your story lacks details and specificity and is inconsistent with your wife’s story, therefore the Finnish Immigration Service does not believe that it reflects a true event.” was coded as both detail and inconsistency.

Interrater reliability analysis

The first and second author coded four cases cooperatively as a learning phase, after which they coded 20% of the sample (n = 11) independently for an interrater reliability analysis. The focus was on identifying assumptions. When evaluating the results, the coders noted a need for stricter and more extensive coding principles for identifying assumptions. The coding principles were therefore revised, and a new set of cases was chosen. In the second interrater reliability analysis, we found that a total of 109 assumptions were identified, of which both coders had identified 88 (80.7%). Of the 21 assumptions that had been identified by only one coder, 8 were excluded after discussion since they did not fit the definition. In approximately half of the inconsistently coded assumptions, we had identified the same assumption, but disagreed on whether it contained one or two assumptions. Following Herlihy et al. (2010), we considered the interrater reliability sufficient.

Revising the sub-themes

After identifying and coding all assumptions, we adapted the sub-themes to better fit our sample. We deleted the sub-theme demeanor since no examples were identified in our sample. Furthermore, we divided the sub-theme assumptions about threats being only temporary into two sub-themes; threat assumed not to be personal and threat assumed not to be durable, as we found that most of the examples related to a threat not being personal, rather than not being persistent. At this point, we observed that assumptions belonging to these two sub-themes were used both in regards to the credibility assessment and the substantiality assessment. The extent to which these statements were psychological assumptions rather than references legal criteria was often unclear in the decisions, and therefore, we decided to include them in the analysis at this stage. Finally, we slightly renamed the remaining themes (for a list of the final themes, see Table 2).

Table 2 Frequencies of the themes and sub-themes of assumptions made by finnish asylum officials.

Themes and sub-themes	n	%	CI % 95
Now: The applicant as a truthful witness	259	57.6	[53.03,62.17]
Sufficiency of detail and specificity	102	22.7	[18.83, 26.57]
Consistency	138	30.7	[26.43, 34.96]
The ability to disclose sensitive information	9	2.0	[0.71, 3.29]
The impact of one assumption on the whole claim	10	2.2	[0.84, 3.56]
There: Plausibility of account	169	37.6	[33.12, 42.08]
Plausibility of the alleged conduct of State and non-State actors	24	5.3	[3.23, 7.37]
Plausibility of the alleged conduct of the applicant	17	3.8	[2.03, 5.57]
Plausibility of the alleged conduct of other actors	7	1.6	[0.44, 2.76]
Threat assumed not to be personal	104	23.1	[19.21, 26.99]
Threat assumed not to be durable	17	3.8	[2.03, 5.57]
Here: The asylum system	21	4.5	[2.58, 6.42]
Delay and disclosure	12	2.7	[1.20, 4.19]
The benefit of the doubt	9	2.0	[0.71, 3.29]

Note. CI = confidence interval. 95% CI were calculated as point estimate +/- 1.96*SE, and SE were calculated as √((p(1 − p)/n).

Next, we thoroughly examined the small number of assumptions (n = 19) which had not been coded into any of the previously described sub-themes. We read the assumptions, their descriptions, and preliminary labels several times and adapted the labels during the process. We identified two new sub-themes to the theme Now: The applicant as a truthful witness that had not been included in the previous studies; the ability to disclose sensitive information and the impact of one assumption on the whole claim. The first sub-theme represented assumptions about what applicants are able and willing to talk about. The second sub-theme reflected instances where the asylum official explicitly stated that the credibility of one element affected the credibility of another element of the claim.

Results

Descriptive results

We identified a total of 449 assumptions, with cases containing 1–22 assumptions. Most assumptions led to negative evaluations or disbelief in the applicants’ accounts (72.4%). Assumptions leading to positive evaluations were found only in the following three sub-themes: consistency (77.4%), detail (14.5%), and the benefit of the doubt (7.5%).

Themes

The frequencies of the themes and sub-themes of the assumptions found in the sample are presented in Table 2. For examples of each sub-theme, see Table 3.

Table 3 Examples of the themes and sub-themes.

Theme	Sub-theme	Example	Assumption
Now: The applicant as a truthful witness	Sufficiency of detail and specificity	“Altogether, your narrative is marked by a lack of subjective specificity. Your claims about the threat from Asaib Ahl Al-Haq is undetailed, and you have given limited answers to the clarifying questions about the threat.” “When you have been asked to disclose more about the attack, your statements have remained narrow and only contained few details that are typical for self-experienced accounts.”	A lack of detail and specificity in applicant’s narratives can be used as indicators of untruthfulness.
Now: The applicant as a truthful witness	Consistency	“You have presented detailed information about the attack on your store, but [name X] and [name Y] statements contain inconsistencies between themselves. [X], you said that you were inside the cafeteria when the men attacked and assaulted you. [Y], you said that [X] had closed the store and was outside when the attack happened. This weakens the credibility of your claims.” “Your identity cannot be confirmed as you have not presented any documentary evidence to support it. However, you have spoken consistently about your background during the process. The Finnish Immigration Service therefore deems your claims about your citizenship, religion, and ethnicity credible.”	Inconsistency within the applicant’s narrative, between two individuals’ narratives, or between external information and the applicant’s narrative can be used as indicators of untruthfulness.
Now: The applicant as a truthful witness	The ability to disclose sensitive information	“When prompted, you have not described your thoughts and feelings regarding the conversion to Christianity [questions X-Y]. You have said that you have chosen the church [name] because the Holy Spirit has touched you, but this experience has been presented very superficially [questions X-Y]. When asked how your life has changed after coming to faith, you have given a very short answer about your life becoming better after having read the Bible [questions X-Y]. The Finnish Immigration Service considers it reasonable to assume that you would be able to describe this important change in your life and the thoughts related to it in more detail, and from a more personally experienced perspective.” “Sexuality is part of an individual’s identity. When one’s identity is in a process of change, there are often several different feelings associated with homosexuality and the recognition of it. You state that your sexual thoughts about men emerged after puberty. However, you have not described the formation of your sexual identity, or thoughts and feelings related to it. […] Your answers regarding your sexual orientation have also been closely tied to having sexual intercourse, not thoughts and feelings you have experienced due to your homosexuality. The Finnish Immigration Service states that there is more to one’s sexual identity than having sex.”	If individuals are unable or unwilling to disclose sensitive personal information, this can be seen as an indicator of untruthfulness.
Now: The applicant as a truthful witness	The impact of one assumption on the whole claim	“[…] you have given wrongful information about your home area, which in itself weakens the credibility of your narrative. Therefore, the Finnish Immigration Service does not accept your claims about the threats from ISIS to be real.” “You have presented photographs and copies of documents which you claim are related to your father’s political activity. Considering the ambiguities and inconsistencies in the other documents you have presented, [these] documents cannot be given much evidential value.”	The perceived credibility of one element can affect the credibility of a later element of the claim.
There: Plausibility of account	Plausibility of the alleged conduct of State or Non-state actors	“Considering the presented Country of Origin Information (COI), your age, story, and lack of special skills and knowledge about weapons, the Finnish Immigration Service does not accept as a fact that you would be subject to a threat of being recruited into armed groups.” “The Finnish Immigration Service finds it implausible that you at the age of 13, after your father’s death, have begun receiving threatening text messages from Mahdi’s army due to your father’s party membership.”	The behavior of a State or non-State actor is deemed implausible and as a consequence, the claim is deemed not credible.
There: Plausibility of account	Plausibility of the alleged conduct of the applicant	“Your narrative contains implausible features. You said that you were unconscious during the rape, but still at times conscious when the rapist told you his name. […] the Finnish Immigration Service does not accept that your rapist was an infamous prisoner wanting to kill you because you recognize him.” “A nurse at the Reception Center has stated that you have started a healthier lifestyle and that you do not consume alcohol at all anymore. […] the Finnish Immigration Service considers it reasonable to assume that a person can stay away from the consumption of alcohol if it causes serious problems. […] Considering all the above, the Finnish Immigration Service does not accept that you would be in danger of serious infringement due to the consumption of alcohol if you were to return to your country of origin.”	The behavior of an applicant is deemed implausible and as a consequence, the claim is deemed not credible.
There: Plausibility of account	Plausibility of the alleged conduct of other actors	“There are several grammatical and typing errors in the documents, which does not support your claim that the documents are from a law firm. […] Considering the previously mentioned facts, the Finnish Immigration Service does not accept your narrative about the events of 2012 to be true.” “Considering your entire narrative, the Finnish Immigration Service finds it implausible that even though [person X], according to your statements was a famous and valued lawyer, he only had one inexperienced supporting lawyer in his office.”	The behavior of another actor is deemed implausible and as a consequence, the claim is deemed not credible.
There: Plausibility of account	Threat assumed not to be personal	“The Finnish Immigration Service concludes that you have been randomly targeted by a shooting due to the country’s general safety situation.” “The previously presented COI supports the fact that marriages between Sunnis and Shias can cause problems on a case-by-case basis. [Person X], you have not presented credible grounds as to why you would be of specific interest for the Shia militia because of your intermarriage or other reasons.	Assumptions about what constitutes a real threat, and which aspects assure that the threat is directed toward the applicant.
There: Plausibility of account	Threat assumed not to be durable	“You have left the area and been away for over one and a half years. Considering the time that has passed since the event and the nature of [it], the Finnish Immigration Service does not accept as a fact that you would be in danger of serious violations from the same persons if you were to return to your home area.” “However, considering that you have not brought forward any other personal reasons than your clan background, and that those violations took place 20 years ago, and that according to current COI, the [X]-minority are not subject to direct violence, the Finnish Immigration Service does not accept that the reason why you originally fled Somalia would put you in danger of violations in the future.”	Assumptions about which threats might subside over time, and how long it takes for them to subside.
Here: The asylum system	Assumptions about delay and disclosure	“You have stated that you have understood the meaning of the hearing, including the obligation to disclose all reasons why you cannot return to your home country, and therefore, The Finnish Immigration Service deems it generally implausible that you have brought these [new] aspects forward only at the appeal stage of the process, when you have not mentioned them at all at the hearings arranged at [place X] and [place Y].” “When you, [person X] and [person Y], arrived in Finland, you told the authorities that you came on a tourist trip to Finland and that you would stay for 42 days. You have only presented your asylum claims when the authorities prevented you from entering to the country. When asked why you did not immediately disclose that you were going to apply for international protection, you said that you were afraid, that you did not have any experience of seeking asylum, and that you had to get to the Reception Center.”	Disclosing new relevant information at a later stage in the process can be seen as an indicator of untruthfulness.
Here: The asylum system	The benefit of the doubt	“Your statements about [ex-husband’s] background was left unspecified. However, considering that you were only 13 years old when you married him and that you only spent roughly one year together with him, the Finnish Immigration Service assumes that you cannot be expected to be able to speak about [your ex-husband’s] background in more detail. The documents that you have delivered from the court corroborates your statements about [ex-husband’s] criminal past.” “When assessing your [daughter’s] danger of being subject to genital mutilation if you and your family were to be returned to your home area, the Finnish Immigration Service applies the benefit of the doubt-principle. […] the Finnish Immigration Service has not been convinced of [parents’] general credibility earlier. However, considering all previously mentioned facts and the best interests of the child, the Finnish Immigration Service does not believe that your [parent X] and [parent Y] lowered credibility is a reason not to apply the benefit of the doubt-principle […]”	Assumptions about when to implement the principle of the benefit of the doubt.

Now: The applicant as a truthful witness

More than half of all identified assumptions related to the applicants’ ways of presenting their claims during the asylum hearings. The four sub-themes represented both the way the applicant presented their claims, as well as what the officials expected the applicants to disclose. The first sub-theme, sufficiency of detail and specificity, consisted of almost a quarter of all assumptions in the total sample. Truthful applicants were expected to provide detailed and specific information about their experiences, including peripheral information about durations, times, frequencies, names, verbatim wordings of conversations and texts, as well as visual or auditive details. A lack of specificity in the accounts was perceived as an indicator of untruthfulness.

The second sub-theme, consistency, was the most frequently occurring type of assumption, making up almost one third of all assumptions. Asylum officials expected truthful statements to be consistent both within and between interviews with the applicant and between family members. Furthermore, applicants’ statements were expected to correspond to external corroborating information, either presented by the applicants themselves or externally retrieved (e.g. Country of Origin-reports collected by the UNHCR).

A small number of assumptions belonged to the third sub-theme, the ability to disclose sensitive information. Asylum seekers were expected to be able and willing to disclose sensitive personal information about, for example, traumatic experiences, personal relationships, sexual orientation, or religious beliefs. An applicant’s inability or unwillingness to describe subjective thoughts and feelings regarding a personal experience was considered as indicative of untruthfulness.

The sub-theme the impact of one assumption on the whole claim, also consisted of a small number of assumptions. In these assumptions, the decision-maker explicitly stated that the perceived credibility of one element affected the credibility of another element of the claim. These assumptions were often related to verbal claims. For example, according to an official, an untruthful claim about the home country also indicated that a claim about a military group was untruthful. Only few assumptions related to other aspects, such as the documents submitted by the applicant: In one case, a document was deemed fabricated, and as a result another unrelated document was also deemed fabricated.

There: Plausibility of account

More than a third of the assumptions belonged to the second theme, relating to plausibility. Three of the sub-themes reflected the plausibility of the behavior of individuals and organizations and two sub-themes reflected what officials assumed personal and durable threats to entail. The first sub-theme, plausibility of the alleged conduct of State or non-State actors consisted of a small number of all assumptions and reflected officials’ beliefs about how State or non-State actors (e.g. paramilitary, or extremist groups) normally function. Decision-makers held various assumptions about the grounds on which military groups base their recruitment. For example, it was not seen as plausible that military groups would have attempted to recruit young applicants or applicants lacking military skills. Officials also held assumptions about when, how actively, and how successfully military groups and States search for and find wanted persons.

A similarly small number of assumptions belonged to the second sub-theme, plausibility of the alleged conduct of the applicant. These assumptions related to the plausibility of the behavior of the applicant, often entailing expectations about how rational individuals behave in so called fear, fight, or flight situations. A variety of different behaviors were deemed implausible by the officials. In one case, it was deemed implausible that an applicant had heard the name of her rapist, since she allegedly had been partially unconscious during the rape. In another case, regarding an applicant with a history of substance abuse, the decision-makers stated that a person can reasonably be assumed to be able to stay away from alcohol if it causes serious problems.

A somewhat smaller number of assumptions belonged to the sub-theme plausibility of the alleged conduct of other actors. These assumptions reflected decision-makers’ expectations about the behavior of other actors, such as family members or relatives. For example, one decision-maker did not believe that a document could have been written by a law firm because of several typing errors. In another case, the official assumed that it was implausible for a renowned lawyer to only have one inexperienced employee.

Approximately one fourth of all assumptions belonged to the sub-theme threat assumed not to be personal. Decision-makers assumed that an applicant was not personally threatened based on a variety of reasons, for example that the threat was related to the general safety situation in the applicant’s home country. In other cases, the applicant was not seen as sufficiently profiled, or the threat was assumed to be directed at a family member or relative and not personally at the applicant. A few assumptions belonged to the related sub-theme threat assumed not to be durable. Decision-makers assumed in some cases that threats would subside over time. However, the time frame of what was considered sufficient time for a threat to pass differed between decision-makers from a few months to several years. Judgements about both of these sub-themes were frequently included in the substantiality assessment, that is, when the claims were compared to the legal criteria. Furthermore, in both these sub-themes, decision-makers frequently used external information (e.g. Country of Origin-reports) to support their decisions.

Here: The asylum system

Less than 5% of all assumptions belonged to the third theme. These assumptions related to the use of the benefit of the doubt-principle, and to the applicants’ knowledge about the asylum procedure and ability to navigate through the asylum system. For the assumptions about delay and disclosure, decision-makers expected applicants to be aware of how and when to apply for international protection, as well as what information and evidence to provide during the processes. Not applying for asylum directly after arriving at the receiving country, or not disclosing all information or evidence promptly at the beginning of the asylum hearing was seen as an indicator of untruthfulness.

In the second sub-theme, the benefit of the doubt, the use of the principle was explicitly stated. Decision-makers gave the benefit of the doubt to applicants both based on personal circumstances (such as young age, or physical and mental health concerns) and contextual circumstances (such as the best interest of a child). Furthermore, sufficient consistency could compensate a lack of detail and vice versa.

Discussion

The current study analyzed assumptions held by first-level decision-makers at the Finnish Immigration Service, expanding on previous research regarding how subjective assumptions held by decision makers affect credibility assessments at all levels of decision-making. Finnish asylum officials seemed to hold similar assumptions to those held by decision-makers at the appeal level, which were the focus of previous studies (Dowd et al., 2018; Herlihy et al., 2010). A notable difference was, however, found regarding assumptions related to demeanor, as no such assumptions were identified in the current sample. It cannot be excluded that demeanor may have had some effect on the officials and their decision making but was simply not written down. Nonetheless, this suggests that credibility assessments in our sample were not based on the behavior or manners of the applicants in the interview situations to any substantial degree, which is in line with the current best-practice guidelines. Demeanor should not be used as an indicator of untruthfulness, as individuals with different background show a variety of behaviors and no reliable behavioral cues to lie detection exist (Bond & Depaulo, 2006; Dowd et al., 2018; Kagan, 2003; UNHCR, 2013).

Now: The applicant as a truthful witness

In many cases, asylum officials expected truthful applicants to present their claims in sufficient detail, with sufficient consistency within and between interviews and between family members. Sufficiency of detail and consistency are described in the current best practice guidelines by the UNHCR as credibility indicators that can be used to distinguish between truthful and untruthful claims. These indicators are, however, partially inconsistent with psychological knowledge. Several researchers and the UNHCR have acknowledged that these indicators should not be used without careful consideration of the naturally occurring inconsistencies and limits of memory, as well as the potential impact of trauma and cultural differences on memory (Cameron, 2010; Cohen, 2001; Herlihy et al., 2012; UNHCR, 2013). Whereas a truthful applicant can be expected to remember the central gist of an event, psychological science does not provide reasons to expect an applicant to remember all peripheral information (Cameron, 2010). Moreover, psychological science does not provide reasons to expect perfect consistency within or between truthful narratives (Herlihy et al., 2012). Trauma, mental health issues, and cultural differences may have an impact on memory, affecting the applicant’s ability to tell a consistent and detailed narrative (Herlihy et al., 2012). In the current sample, Finnish decision-makers often seemed to lack knowledge about how naturally occurring variation in memory might affect the narratives of asylum seekers. In a few exceptions, officials stated that they accepted a statement as truthful even if it was, for example, undetailed or inconsistent, and this was explained by factors such as the age or mental health of the applicant. Nevertheless, alternative explanations for inconsistent or undetailed statements were rarely considered.

Moreover, we found assumptions reflecting the expectation that truthful applicants can and are willing to provide detailed information about sensitive personal experiences such as the development of one’s sexual identity and religious conviction. Studies have found that disclosure of sensitive personal information in asylum hearings is considered difficult by many asylum applicants, and that disclosure is related to various individual and contextual factors such as trauma, stigma, gender, educational background, and interviewer characteristics (Baillot et al., 2014; Bögner et al., 2010; UNHCR, 2013). Furthermore, cultural differences may impact the content and focus of memories. The collectivistic, general memories focusing on social interaction that asylum seekers may present, might not fit the Western asylum official’s expectation of self-focused and detailed memories (Wang, 2016). A person from a non-western culture might, thus, not have encoded detailed, self-focused memories about thoughts and feelings regarding, for example, one’s sexual development. Various factors related to intercultural communication may also affect disclosure, such as the perceived power distance, or different understandings of what details are considered relevant (Granhag et al., 2017; Hope et al., 2022). To automatically dismiss unwillingness to disclose sensitive information as an indicator of untruthfulness is not in line with psychological science. Finnish asylum officials rarely seemed to—at least in writing—consider possible alternative explanations to why an asylum seeker failed to disclose personal experiences.

Finally, in the written determination letters, a perceived lack of credibility in one statement could undermine the credibility in later statements. This theme is not related directly to empirical psychological knowledge, and thus not a focus of this review. Nevertheless, it can be noted that not processing each element of an applicant’s claim objectively and impartially is not in line with the recommendations of conducting credibility assessments (Qualification Directive, The European Parliament, 2011).

There: Plausibility of account

Asylum officials were found to judge a variety of behaviors, actions, or lack of actions, as implausible or unlikely both generally and in specific situations. Many assumptions seemed to be based on a subjective view of how a reasonable person (asylum seeker or persecutor) would act. Notably, officials seldom explained their reasoning behind decisions, nor did they explore possible alternative explanations for unlikely behaviors. Assumptions about plausibility are problematic since human behavior is highly variable and unpredictable, and it is not possible to predict with certainty how individuals will react and behave in specific situations (Granhag et al., 2017; Reifels et al., 2013). Moreover, according to Sweeney (2007) comparing the behavior of a persecutor to a standard of how a reasonable person would have acted effectively means judging if and how the persecution could have been executed more effectively. In other legal contexts, a standard of what can be expected of a reasonable person in terms of liability may be valid, however, the benefit and appropriateness of applying a similar standard to persecutors in foreign countries is questionable (Sweeney, 2007).

Furthermore, the definition of plausibility lacks clarity and consensus. The UNHCR highlights that plausibility, although described as a credibility indicator, should be used with considerable caution and never be the sole ground for rejecting asylum (UNHCR, 2013). The UNCHR recommends defining plausibility as the “realistic likelihood of events”, commonly interpreted as implying that a statement is implausible if a decision-maker deems it to be “beyond human experience of possible occurrence” (Maegherman et al., 2018). Although the definition is precise, according to whom something is implausible remains a subjective element. In other words, despite the asylum process being required by law to be objective and impartial, decision makers must evaluate what is realistic based on their subjective views (Herlihy & Turner, 2009; Maegherman et al., 2018). Our findings highlight this problematic aspect of judging plausibility.

Our findings support the notion made by previous authors that judgements about plausibility in asylum decisions is based on personal first- or second-hand commonsense judgements, speculations, stereotypes, and subjective perceptions of risk (Herlihy & Turner, 2009; UNHCR, 2013). These judgements fail to consider possible differences in norms, customs and behaviors between cultures and societies or the individuals’ personal circumstances. Moreover, shortcuts in our thinking processes and a range of biases at different levels affect all decision-making (Dror, 2020). In contexts where it is impossible to provide feedback on accuracy, decision-makers can become more and more biased. This risk can be reduced only with continuous training on how to counteract bias (Dror, 2020; Pompedda et al., 2017). The UNHCR (2013) concludes that if plausibility is used as a credibility indicator, the assessment should be based on independent, objective, reliable, and time-appropriate evidence, and be thoroughly described with clear references to the evidence. Finnish asylum officials do not seem to follow this recommendation, as the explanations behind the reasoning found in our sample were limited and unstructured.

Determining whether threats are genuine and durable is an important part of evaluating the legal grounds for granting asylum. However, in the current sample, these statements were often difficult to separate from statements about the credibility of an applicant’s claims. Additionally, the assumptions regarding whether threats are genuine, and durable were often intertwined with external information. These factors made it difficult to interpret to what extent they were related to a lack of consideration of psychological knowledge. We, nevertheless, observed a large variety in these assumptions, especially in relation to what durable threats entail. These types of assumptions might result in inconsistent decision-making by asylum officials, and consequently, to legally uncertain determinations.

Here: The asylum system

Although rare, we found asylum officials to hold assumptions about the use of the benefit of the doubt-principle, and about applicants’ abilities to understand and navigate through the asylum system. Both aspects are based on legal directives and, therefore, relate more to judicial issues than to psychological science. According to the Qualification Directive, The European Parliament (2011), an applicant must apply for asylum, present all relevant information, and submit all available evidence as early as possible. Nevertheless, there could be several alternative explanations for why applicants might not act according to these principles. First, particularly in the beginning of or during the first hearing, the applicant may be stressed by the situation. It is well established that stress during the interview hampers memory (Shields et al., 2017). Second, due to the constructive nature of memory the applicant may remember additional information at a later recall and add relevant details to the claim at a later stage of the procedure. Third, applicants suffering from mental health issues may not be able to keep track of their obligations as asylum seekers, or which aspects of the claim they have not disclosed yet. Fourth, difficulties due to intercultural communication might create situations in which the applicant does not understand what constitutes relevant information or what “tell everything” entails. Fifth, differences in power distance might make the applicant hesitant to present objections against or additions to the official. Sixth, applicants may find it difficult to disclose sensitive information to officials and interpreters if they doubt the confidentiality of the interview, or distrust officials all together (Herlihy & Turner, 2009; Kälin, 1986). Finally, trauma, guilt, shame, stigma, distress, or nervousness might affect the disclosure of information negatively. In sum, as Coffey argues: delayed disclosure may be “entirely consistent with the reality of the event” (Coffey, 2003, p. 415). Officials should carefully consider such alternative explanations when assessing these issues.

In our sample, the benefit of the doubt was used sparsely. Asylum officials might not, however, always explicitly state when they use the principle in the written determination. Additionally, the UNHCR (2013) has found that officials are often uncertain of how to apply the benefit of the doubt. Nevertheless, in the examples we identified in our sample the benefit of the doubt seemed to be applied appropriately, according to international standards. Such examples of good practice could be collected by researchers and used in training and guidebooks for the officials carrying out the indeed very complex and difficult task of conducting asylum interviews and/or making decisions based on these interviews.

Limitations

Although our coding principles were based on previous findings and we worked to establish reliable coding principles and definitions, the coding of assumptions was complex and may have been affected by our subjective expectations and own assumptions.

Differences in the written determinations between individual officials were ample. The style of describing and explaining aspects regarding the credibility of an applicant varied to a large extent. Consequently, it was sometimes challenging to understand which statements were related to which conclusions. It is possible that the officials did consider alternative explanations for negative credibility findings in their decisions but did not write them down. The coding was further complicated by the fact that the determination letters did not include an explicit section for the credibility assessment, which made it difficult to separate credibility findings from the substantiality assessment.

Moreover, the sample of determinations was relatively small and consisted mostly of rejected cases with negative assumptions. This is because Finnish officials are not obligated to document their reasoning for accepted cases, which makes written explanations of positive credibility findings rare. Only documenting and analyzing the decision-making underlying negative cases may create serious distortions in both research and asylum decision-making. Additionally, we included even infrequent sub-themes since they have been included in previous studies and give a more detailed insight into this rather understudied field of research.

Recommendations

Officials at the Finnish Immigration Service were found to hold a number of assumptions that are not in line, or only partially in line, with empirical research on human behaviour and memory. Our findings give cause for recommendations for future practice and research necessary to improve the quality of the decision-making processes of asylum officials and protect the integrity of the asylum procedures.

Firstly, asylum officials should be provided with more training on how memory functions and on factors affecting communication and decision-making. Persecution is a traumatic experience, and, therefore, stress and the possible impact of trauma and trauma-related mental health disorders must be well understood by asylum officials. Improved knowledge about cultural differences and challenges related to intercultural communication is also vital for asylum officials whose entire work is carried out in an intercultural context. Understanding and acknowledging different cognitive biases that affect decision-making, especially in situations where ground truth is difficult to establish, is equally important. Guidelines and handbooks already incorporate these aspects to some extent, but these aspects should also be incorporated in the practical training that officials receive. Research from other legal contexts (e.g. child interviewing) has found that extensive training, continuous feedback, and follow-up sessions are needed to successfully implement new practices (Powell, 2008). Similarly, authors have listed a number of steps of how to overcome bias in and improve legal decision-making (Burke, 2006). Developing and extending the training that asylum officials receive has been recommended (see, e.g. Schoenholtz et al., 2007), and would likely improve the quality of the asylum procedures.

Second, requiring asylum officials to structure and explain their reasoning behind all credibility findings in a more explicit manner would increase the transparency of the decision-making process. Including a separate section for describing only credibility aspects would improve the overall structure of the determination letters. Decision-makers should explicitly write down all arguments with clear references to the evidence and systematically seek alternative explanations to all findings by actively considering the opposite (e.g. Hirt & Markman, 1995; O’Brien, 2009). Other debiasing methods worth exploring include alternative stories or systematic tools such as checklists (Schmittat et al., 2022; Sibbald et al., 2013). These measures could minimize the effect of bias and subjective assumptions on the decision-making. Additionally, this would also make it possible for the official, the applicant, as well as lawyers, judges, and policy makers to understand and evaluate the reasoning on which the decision has been made. Officials in Finland, as well as in many other countries, are currently not obligated to address the credibility assessment in granted asylum cases. Addressing the credibility assessment and explaining the reasoning behind credibility findings in all cases, regardless of outcome, would further increase the transparency.

Conclusions

Our findings show that Finnish asylum officials to a substantial extent base their decisions on assumptions that are not grounded in psychological science. Using UNHCR’s credibility indicators correctly requires careful consideration and caution by asylum officials. Our study further confirms the need for policy makers, case workers, and interdisciplinary experts to improve the validity of the current methods to protect the integrity of the asylum process. Importantly, this should entail better training in psychology, in particular, investigative interviewing, memory and decision making for asylum officials, requiring officials to increase the transparency in their reasoning, and systematically combating bias in decision-making.

Availability of data

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by a grant from Waldemar von Frenckell’s Foundation to the first author (J.S), a grant (decision number: 151173) from The Swedish Cultural Foundation in Finland to the third author (J.A), and a grant from Sundell’s Foundation to the last author (J.K).