Stakeholder Legitimization of the Provision of Emergency Centralized Accommodations to Displaced Persons

: Sudden population inﬂuxes into cities—such as those seen during post-disaster migration—place unexpected demands on the urban housing system. Decisions made during these inﬂuxes are often controversial, potentially hindering the ability of the organizations involved to respond. This study’s objective was to explore strategies (e.g., types of information shared and types of accommodation chosen) that can be used during decision-making processes when providing emergency accommodations to increase stakeholder acceptance, and thus lead to sustainable institutional responses. This study speciﬁcally sought to shed light on how, during the Refugee Crisis in Germany of 2015 and 2016, stakeholders legitimized decisions made to provide centralized emergency accommodations to displaced persons. Making this study possible were 25 semi-structured interviews with utility, government, nonproﬁt, and company employees involved in the provision of centralized accommodations for displaced persons. Interviews were conducted in 2016 and underwent a qualitative analysis. Results indicate that stakeholders primarily legitimized the provision of centralized accommodations based on convictions of right and wrong (moral legitimacy), while they legitimized decisions to not provide such accommodations based on their understanding and experience of practical barriers (cultural-cognitive legitimacy). Recommendations arising from this study include the following: provide information to stakeholders about accommodations’ livability (to gain consequential legitimacy) and past successes (to gain comprehensibility legitimacy), adapt regulations to help stakeholders use procedural legitimacy, and prefer fully renovated buildings or modular housing to buildings with no major renovations or container housing (to gain consequential rather than procedural legitimacy).


Introduction
By the end of 2018, the global population of individuals forcibly displaced was 70.8 million [1]. Recently increasing at a faster pace, that number is twice what it was in 1998 [2]. Between 2003 and 2008, for instance, the number of persons newly displaced each year never exceeded 5.9 million; every year between 2013 and 2018, this number was 10.3 million [1,3]. This increase is largely attributed to ongoing conflicts and deteriorating political situations, with 67% of all refugees worldwide originating from five politically unstable nations-the Syrian Arab Republic, Afghanistan, South Sudan, Myanmar, and Somalia [1]. Given this trend and the worsening of political tensions worldwide, the amplitude legitimize both the provision and non-provision of those centralized accommodations. We then provide an overview of the types of accommodations that were used during the Refugee Crisis along with the corresponding stakeholders' perspectives. The perspectives are summarized based on data from interviews and select legitimations.
The relevance of these research questions emerged from previous work performed by Faure et al. [12]. Germany was selected as a strategic location, since over one-third of the 2015 European Union asylum applicants registered there [13]. Additionally, German Chancellor Angela Merkel made decisions on German migration policies that were both highly criticized and applauded by the German public and the rest of the world (as highlighted in the media [14][15][16]).

Overview of the Process of Accommodation of Displaced Persons in Germany
During the influx of displaced persons, Germany altered its process of accommodating displaced persons from that process during non-emergency times (see Figure 1). It should be noted that Germany required asylum seekers to stay for a minimum of three months in their initial reception facilities before being allowed to move into private apartments [17]. Asylum seekers struggled to find private apartments, however, thanks to housing shortages, discriminatory housing markets, restrictive regulations, and their inability to find work [18]. Refugees who were granted asylum also faced these problems due to the challenge of finding jobs, owing to language issues or nonrecognition of their diplomas. As a result, asylum seekers (and refugees) tended to remain in centralized accommodations throughout the entirety of the asylum procedure-beyond the minimum three months required-even after being granted asylum [19]. Table 1 provides an overview of the types of accommodations commonly used as initial receptions facilities, collective accommodations, and emergency accommodations during the Refugee Crisis. Some emergency accommodations were temporary, such as sport halls that required recommissioning before local schools could return to using them. Many of them, though, were renovated for long-term use as initial reception facilities or collective accommodations. There was no clear technical delineation between emergency accommodations and other centralized accommodations for displaced persons. Some government agencies, for example, considered container housing to be short-term solutions (e.g., two or three months), while other agencies considered them to be long-term (e.g., five or more years).   [17,20]). Table 1 provides an overview of the types of accommodations commonly used as initial receptions facilities, collective accommodations, and emergency accommodations during the Refugee Crisis. Some emergency accommodations were temporary, such as sport halls that required recommissioning before local schools could return to using them. Many of them, though, were renovated for long-term use as initial reception facilities or collective accommodations. There was no clear technical delineation Sustainability 2020, 12, 284 4 of 26 between emergency accommodations and other centralized accommodations for displaced persons. Some government agencies, for example, considered container housing to be short-term solutions (e.g., two or three months), while other agencies considered them to be long-term (e.g., five or more years). Table 1. Types of accommodations typically provided to displaced persons during the Refugee Crisis in Germany.

Accommodation Denomination Types of Accommodations Provided to Displaced Persons
Initial reception facilities and collective accommodations [20] Various types of buildings owned or rented by the government, including: • buildings entirely or partly renovated, such as former office buildings, schools, or factories; • buildings specifically built to host displaced persons; • container housing (assembled container units); • modular housing made of standard construction units (e.g., standard wall surfaces).
Emergency accommodations [20] Short-term solutions-intended for a few months-to prevent displaced persons from being homeless in Germany. They typically were: • sport halls in local schools that could accommodate up to 1000 displaced persons; • light-frame buildings (e.g., tents, inflatable domes); • former schools, airports or office buildings, where only minor renovations were undertaken prior to hosting displaced persons.
Responsible for accommodating displaced persons were state and city-level government agencies. When they went about identifying locations (e.g., existing buildings or empty land), government agencies may have been advised by different organizations (e.g., chambers of architects) and may have collaborated with private companies. After identifying feasible locations, architects and companies were contracted by the government agencies to renovate, design, or construct buildings. Following this, nonprofits and companies were contracted to manage those accommodations and provide daily services to displaced persons, while the maintenance work was contracted (and monitored) by government agencies. In this study, centralized accommodation refers to initial reception facilities, collective accommodations, or emergency accommodations.
In 2015 and 2016, Germany had to expand its timeframe for providing housing for displaced persons due to the sudden influx of displaced persons. Measures to reduce the timeframe included simplifying several permitting processes and removing the requirement for architecture competitions to select agencies responsible for the design of accommodations.

Accommodations During Emergency Situations
Previous research regarding accommodations during emergency situations primarily focused on three areas-refugee camps in developing countries, internal displacements due to natural disasters, and decentralized accommodations for internationally displaced persons. Previous studies have focused on refugee camps for both internally and internationally displaced persons in developing countries. Some of these studies have focused on the way institutions plan these camps (e.g., [21,22]), but most of the literature focuses on physical and mental health of those residing. Guthmann et al., for example, studied the effects of inefficient water and sanitation services [23]. Toole and Waldman studied the public health aspects of refugee situations [24]. However, the assessment of camps for displaced persons in developing countries does not address the impact of the accommodations in emergency situations on the hosting city's infrastructure system. Other research topics include natural Sustainability 2020, 12, 284 5 of 26 disaster-related internal displacements in both developing and developed countries (e.g., [25][26][27][28]). Previous research regarding disaster-related displacements typically pairs emergency responses with sustainable recoveries (e.g., [29][30][31][32]). The information sought in this study complements this existing knowledge; the international displacements (mostly from the Middle East) and subsequent emergency response (in Germany) is geographically distinct from the recovery established in displaced persons' countries of origin. Additional literature focuses on the long-term, decentralized accommodations for internationally displaced persons (e.g., [33][34][35]); however, the time scale of those cities' responses is 3 to 10 years, reflecting the time needed to provide displaced persons a stable housing situation (e.g., private flats). Presently, there is limited knowledge regarding centralized accommodations for internationally displaced persons in developed countries and the impact of this rapid population influx with limited front-end planning on centralized accommodations. To achieve sustainability, decision-makers must give attention to planning such accommodations in times of emergency. This fact was made evident by Bris and Bendito and their study of Spain during the European Refugee Crisis [36]. This study aims to address this gap in knowledge by providing insight into different institutional responses to a sudden high influx of displaced persons in a developed country.
Stakeholders involved in the emergency accommodation of displaced persons are diverse. In the context of refugee camps, the following stakeholders were identified by Xu et al. [37] using a case study of a Syrian refugee camp: the United Nation High Commissioner for Refugees (UNHCR), nonprofits, host governments, community organizers, local communities, and refugees themselves. Refugee camps are defined here as "temporary transitional population centers" that can evolve to a "more complex and stable ecosystem [when] protracted" [37]. In developed and politically stable countries, governmental, nongovernmental, and refugee organizations were also identified as a system of major stakeholders [38]. In addition to these stakeholders, and in the context of the provision of centralized housing to displaced persons within cities (outside of refugee camps), architecture and civil engineering companies were are also identified as key stakeholders [39].

Legitimacy Theory
The theoretical basis of this study is predicated on the intuition that emergency response situations are strongly influenced by stakeholders' desire to legitimate action. Emergency responses often lack guidelines or regulations on how to respond to sudden disruptions. As such, individuals involved in emergency responses may try to react according to their own appreciation of the situation. "Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" Suchman [11]. Legitimacy was studied in different disciplines such as psychology (e.g., [40]) and political studies (e.g., [41]), which mainly define legitimacy by focusing on mechanisms of authority. In this paper, however, we use a definition of legitimacy from organization studies (as defined by Suchman [11]), which more broadly focuses on legitimation mechanisms within institutions. Applying such a theoretical framing allows us to understand how and why decisions are made in all types of institutions involved in providing centralized accommodations to displaced persons.
Legitimacy can be used to explain decision-making processes as it directly influences decision makers (e.g., CEOs, managers), and other individuals involved in such processes within institutions. Notably, power is not always a top-down process, and legitimacy within institutions can result in a bottom-up process [8]. Power "can arise out of mobilization of subordinate groups as they attempt to advance their own values and interests" [8] (page 73). With this in mind, how individuals within an organization use legitimacy within a decision-making process should be considered as they set forth organizational goals (e.g., selecting accommodation types). As discussed by Dowling and Pfeffer, "Legitimacy and social norms and values constrain the actions taken by individual organizations" [42] (page 131). For example, legitimacy can negatively affect productivity in collaborations [43], which are necessary when providing centralized accommodations to displaced persons due to the numerous entities involved. Similarly, Beaven et al. found that legitimacy contributed to emergency responses Sustainability 2020, 12, 284 6 of 26 after natural disasters [44]. Looking at centralized accommodations through this legitimacy lens could aid in decision-making to ensure effective adaptation of urban housing systems to diverse, rapid population influxes. Results can identify the types of centralized accommodation solutions preferred by stakeholders, based on their personal experiences, beliefs, and interests. Further, if decision makers recognize the types of centralized accommodations that will (or will not) be accepted by relevant institutions, they might be able to streamline their accommodation strategies by choosing the most widely supported options. Finally, understanding how centralized accommodations are legitimized allows decision makers to justify their choices for a better social acceptance amongst involved institutions.

Materials and Methods
This paper is an exploratory work that empirically builds on 25 ethnographic, semi-structured interviews with multiple types of stakeholders involved in the provision of centralized housing to displaced persons during the Refugee Crisis. Ethnographic, semi-structured interviews "provide complex textual descriptions of how people experience a given research issue" through the collection of personal histories, experience, and perspectives [45]. A qualitative analysis of these interviews was used to identify trends and emergent themes in decision-making processes in a context about which existing information is scarce [46,47]. Similar methods of qualitative analysis of interviews were used in previous studies and were proven successful in identifying patterns in stakeholder perspectives (e.g., [48][49][50]; these studies used between seven and 15 semi-structured interviews). More specifically, qualitative coding was used to capture interviews' "primary content[s] and essence" [51] (page 4) and "retrieve and categorize data that are similar in meaning" [52]. It should be noted that the qualitative analysis of interviews is complimentary to other methods that assess decision-making processes. For instance, to obtain a more representative distribution of stakeholders, data could be gathered using surveys (e.g., [53,54]), social media (e.g., [55]), or workshops (e.g., [54]). Complimentary methods used to assess decision-making procedures include Analytic Hierarchy Process [56] and participatory mapping [54].
To conduct the ethnographic interviews, investigators followed guidelines set by Spradley [51]. Topics covered during interviews included the following: current position and responsibilities; design, construction, and renovation of centralized accommodations; government and other organizations' responses to the Refugee Crisis; and the collaboration between stakeholders during the provision of housing during this period. Questions asked during interviews were a combination of open questions and follow-up questions. Most interviews were prepared and conducted by two investigators. Using more than one investigator has been found to strengthen a study, enhance its creativity, and bolster confidence in the findings [57]. The investigators were of different nationalities (American and French), bringing complementary insights during the interview and analysis processes.
For this study, the research team carried out interviews during the summer of 2016. Snowball sampling was used to ensure relevant participant knowledge and expertise. Further, participants were recruited based on the criteria suggested by Spradley for selecting good informants for ethnographic interviews [51]. Informants were drawn from a broad range of stakeholders involved in the project of providing emergency centralized accommodations for displaced persons in 2015 and 2016. Interviews were conducted across four cities: 11 from City A, two from City B, five from City C, and seven from City D. The cities' total populations spanned 500,000 to 3.7 million in 2016. Cities were amongst the receivers of the country's largest numbers of asylum applications in 2015: that year, those cities received between 4230 and 54,324 (for the largest city) asylum applications. While these cities spanned a wide political spectrum, they all reacted to the large influx of displaced persons by purposely building and renovating centralized accommodations for displaced persons.
All informants were at least 25 years old and had held their current positions for more than six months. Their backgrounds spanned multiple roles and positions (see Table 2), as we wanted to capture the perspectives of all those involved. Four informants were women and 21 were men. "Utility," in this paper, refers to a water and sanitation utility. The two architects involved in urban planning (Table 2) were contracted by two cities to perform feasibility studies for temporary housing. Amongst the seven remaining architects (Table 2), three were commissioned by private clients to design modular and container housing, and four were commissioned by cities to design the renovations of office buildings (for two of them) and other buildings such as schools (for two of them). Amongst the four informants working in other companies (Table 2), two CEOs of construction companies were interviewed: one company was building two centralized accommodations for displaced persons in 2016, and the other one was involved in the advising of the city regarding population dynamics. Another informant was working in a marketing company to advise the local government about the design of centralized accommodations. One informant was working in a real estate company commissioned by the city, acting as a portfolio manager for centralized accommodation locations. The three informants working in nonprofits (Table 2) were the following: (1) a manager of an emergency shelter (a former office building), who was also responsible for managing renovations to transition to a more permanent shelter, (2) an informant responsible for selecting potential centralized accommodation locations, and (3) an informant involved as a social worker in the planning of minor renovations in an existing shelter. Six of the informants working in government agencies ( Table 2) were responsible for making final decisions about portfolios of centralized accommodations in three of the studied cities, and two were responsible for permitting such accommodations. The informant working in a utility ( Table 2) was involved in advising the city about the construction work within centralized accommodations to provide water and sanitation services. Overall, the informants were involved in all the different phases of the making of decisions about the provision of centralized accommodations to displace persons during the Refugee Crisis. These phases included: the selection of locations, the permitting of accommodations, and the design and implementation of renovations or new construction of accommodations (before and after displaced persons' arrival). Interviewees spanned a wide range of stakeholders involved in these phases, given the multiplicity of cities, types organizations they worked in, and roles.
When needed, a German interpreter was present to overcome language or cultural barriers. With permission, 22 of the 25 interviews were audio recorded, providing more than 20 hours of audiotape, which were subsequently translated to English, as needed, and transcribed. During the three interviews that were not audio recorded, investigators took detailed notes.
The interview content was coded for excerpts (de)legitimizing actions made by different entities to provide centralized accommodations to displaced persons. Excerpts of delegitimizing actions consisted of interview content that attributed legitimacy to the choice made by entities to not take a specific action. Codes applied correspond to legitimacy types and subtypes, as defined by Suchman [11] and are summarized in Appendix A, Table A1. While legitimacy has been researched extensively in organization studies (e.g., [58]), the authors have decided to use Suchman's widely accepted typology. This typology is currently perceived as a valid framework used to asses organizational legitimacy ( [59]) and is the theoretical point of departure of many recent studies in this field (e.g., [59][60][61][62]). In fact, using Suchman's typology reduces the conceptual ambiguity associated with legitimacy, as noted by Suddaby et al.-"Both the importance and the conceptual ambiguity of legitimacy have rather increased than decreased in recent years, making legitimacy a central and widely used but often confusing concept in management research" [59].
An example of coding for legitimacy types is as follows: an architect was asked if he agreed with the decisions made by the city's government to finance the creation of a new centralized accommodation. The informant replied: "Mostly; it's the newest building in this area and it upscales maybe the area." This excerpt was coded as pragmatic legitimacy, as the informant, in justifying the new shelter, was anticipating it having a positive effect on the city, of which the informant was part. More precisely, this excerpt was coded to influence legitimacy since the informant was focusing on benefits provided to a large entity-i.e., the city.
Interview content was coded using the software Dedoose [63]. Codes for this analysis were defined using a coding dictionary [64] that was iteratively developed [65]. Definitions, select examples, and key words to look for were included in this coding dictionary. The legitimacy coding methodology is summarized in Appendix B: Figure A1. provides a flow chart of the overall coding process and Figures A2-A4 provide flow charts of the coding process for the specific types of legitimacy (pragmatic, moral, and cognitive). The coding dictionary was verified through interrater reliability checks to ensure coding replicability [66]. Every code application was verified with codes being deleted in cases of disagreement between the investigators, resulting in a kappa value of 1.0. Each excerpt coded corresponds to one specific idea or argument developed by informants during interviews. For example, an informant was asked about renovations that were required on centralized accommodations. He replied: "This is not my responsibility, this is the [governmental agency's] responsibility." Two excerpts were coded, since the first part delegitimized the informant's involvement, while the second part legitimized the government's involvement.
After the legitimacy coding, coded excerpts underwent a secondary analysis to identify emerging themes in the following categories: (1) reasons explicitly cited to (de)legitimizing the provision of centralized accommodations for displaced persons, and (2) types of accommodations specifically legitimized.
While informants had multiple backgrounds (e.g., cities and roles), the trends in the use of legitimacy and in the results of the secondary analysis were recurrent. Additionally, the coding frequencies were used in this study as indicators of key topics, and authors complemented these results by reporting recurring themes and quotes from multiple interviews (e.g., [67]). Doing so increases the transparency and validity of results and their interpretation.

Stakeholder (De)legitimization of Providing Centralized Accommodations to Displaced Persons
Informants (de)legitimized actions taken by different entities to provide centralized accommodations to displaced persons. Those entities discussed included local or national government agencies, the German people, local communities, nonprofits, companies, individual stakeholders, displaced persons, informants themselves, and an entity formed by all stakeholders. For instance, an informant said, "I think thanks to [centralized accommodations] we won't have the situation next winter that people have to freeze outside." In this case, the informant was legitimizing the actions of all stakeholders who worked towards the provision of centralized accommodations for displaced persons. When the use of legitimacy was part of their argument (e.g., by emphasizing that an action is the right thing to do), informants used legitimacy intentionally, but they also used it unintentionally when expressing only their opinion. In this section, we give an overview of the reasons they cited directly as relevant to justifying their perspectives about centralized accommodations, and of the different types of legitimacy they used to do so. Figure 2 shows the frequency with which the different reasons that emerged during the secondary analysis were explicitly mentioned by informants to respectively (de)legitimize the actions taken to provide centralized accommodations to displaced persons. A total of 383 excerpts were coded as reasons cited to legitimize actions, and 77 excerpts were coded as reasons cited to delegitimize actions. The reason most often cited to legitimize the provision of centralized accommodations (47% of excerpts) corresponds to a willingness to improve the livability (living conditions) of accommodations, both locally and at the country level. The livability of accommodations includes overall condition, available space per person, privacy, and safety (e.g., fire safety) within those accommodations. To highlight the need for renovations or the need for new accommodations, for example, 24 out of the 25 informants discussed the poor livability of select centralized accommodations, such as lack of privacy. Notably, the livability of accommodations also comprised 21% of coded excerpts delegitimizing the provision of centralized accommodations. This is partly due to seven informants delegitimizing the construction of new collective accommodations by emphasizing the need for more immediate actions to prevent displaced persons from being homeless. These informants indicated that planned collective accommodations would be set up after several months while emergency solutions should be found within a few days.

Reasons Explicitly Mentioned
The second most frequently cited reason for legitimizing the provision of centralized accommodations were the regulations ( Figure 2). Informants typically referred to existing federal and state requirements for minimum living standards in displaced persons' accommodations, and regulations citing organizations (e.g., utilities, government agencies) responsible for different steps of the accommodation project. Interestingly, the second most frequently cited reason cited for delegitimizing the provision of accommodation was the existence of regulations (30% of the coded excerpts). Additionally, validation using other persons' perspectives was the second most frequently cited reason used for delegitimizing this project; however, it was the least recurrent reason used to legitimize this project.
The integration of displaced persons represents only 12% of the excerpts for legitimizing the provision of centralized accommodations, but informants, when discussing it, seemed strongly convinced of its benefits when it came to integrating displaced persons into the city. One informant said that the way centralized accommodations are distributed throughout the city is directly linked to successful integration of displaced persons: "[T]his can also be an issue if the refugees are in the neighborhoods far from the city center because I think in the city center is very good; this is very easy to integrate the people". The reason most often cited to legitimize the provision of centralized accommodations (47% of excerpts) corresponds to a willingness to improve the livability (living conditions) of accommodations, both locally and at the country level. The livability of accommodations includes overall condition, available space per person, privacy, and safety (e.g., fire safety) within those accommodations. To highlight the need for renovations or the need for new accommodations, for example, 24 out of the 25 informants discussed the poor livability of select centralized accommodations, such as lack of privacy. Notably, the livability of accommodations also comprised 21% of coded excerpts delegitimizing the provision of centralized accommodations. This is partly due to seven informants delegitimizing the construction of new collective accommodations by emphasizing the need for more immediate actions to prevent displaced persons from being homeless. These informants indicated that planned collective accommodations would be set up after several months while emergency solutions should be found within a few days.
The second most frequently cited reason for legitimizing the provision of centralized accommodations were the regulations (Figure 2). Informants typically referred to existing federal and state requirements for minimum living standards in displaced persons' accommodations, and regulations citing organizations (e.g., utilities, government agencies) responsible for different steps of the accommodation project. Interestingly, the second most frequently cited reason cited for delegitimizing the provision of accommodation was the existence of regulations (30% of the coded excerpts). Additionally, validation using other persons' perspectives was the second most frequently cited reason used for delegitimizing this project; however, it was the least recurrent reason used to legitimize this project.
The integration of displaced persons represents only 12% of the excerpts for legitimizing the provision of centralized accommodations, but informants, when discussing it, seemed strongly convinced of its benefits when it came to integrating displaced persons into the city. One informant said that the way centralized accommodations are distributed throughout the city is directly linked to successful integration of displaced persons: "[T]his can also be an issue if the refugees are in the neighborhoods far from the city center because I think in the city center is very good; this is very easy to integrate the people".
Finally, as the cities in which the study was conducted were growing cities, the overall population growth was also discussed by informants, who primarily referenced it to legitimize the provision of housing. Indeed, 10 informants discussed population growth related to displaced persons as it related to the overall population growth of the city, highlighting the notion that, regardless of the Refugee Crisis, new accommodations were needed.

Legitimacy Types
In total, 841 excerpts were coded legitimizing the provision of centralized accommodations for displaced persons, while 216 excerpts were coded delegitimizing centralized accommodations. Informants, thus, legitimized this project 3.9 times more often than they delegitimized it. Figures 3 and 4 illustrate the frequencies at which informants used the three types of legitimacy and nine subtypes of legitimacy to respectively legitimize and delegitimize the provision of centralized accommodations to displaced persons.
The results shown in Figures 3 and 4 suggest that informants were more likely to use a normative evaluation (i.e., moral legitimacy) of stakeholders' actions to legitimize the provision of centralized accommodations than to delegitimize it. Otherwise stated, informants held a conviction that "the right thing to do" was to accommodate displaced persons, as opposed to not providing accommodations. The results also indicate that the portion of pragmatic legitimacy, which relies on foreseen benefits, was higher in delegitimizing excerpts than in legitimizing excerpts. The foreseen benefits can be direct benefits to informants (e.g., a job opportunity or their salary) but also indirect benefits (e.g., benefits to the city). Figure 3 shows that the most frequently cited legitimacy subtypes used to legitimize the project are procedural, comprehensibility, and consequential. Taken together, these account for 60.5% of legitimizing excerpts.
When using procedural legitimacy, informants thought that "the right thing to do" was to do their best and follow procedures they thought applicable, independently from the results of those procedures. For example, three informants justified select actions by highlighting that those actions were "how they do it in Germany." Similarly, an informant legitimized his agency's decision to improve fire safety in some accommodations by saying, "Fire protection is a big thing for us in [our city]. That was really important for us." The informant was, thus, focusing on the procedure that she thought appropriate (since in line with her city's values) rather than on its outcome. Informants primarily used procedural legitimacy when they mentioned existing German regulations and the livability of centralized accommodations. These justifications accounted for 42% and 31% of reasons, respectively. Interestingly, most regulations mentioned were related to the livability of centralized accommodations. For example, an informant legitimized the involvement of the government agency she was working for by saying: "It's our job to make sure that these shelters are in hygienically perfect condition. That's the same for hospitals. We're there regularly to control the hygiene".
In total, 841 excerpts were coded legitimizing the provision of centralized accommodations for displaced persons, while 216 excerpts were coded delegitimizing centralized accommodations. Informants, thus, legitimized this project 3.9 times more often than they delegitimized it. Figures 3  and 4 illustrate the frequencies at which informants used the three types of legitimacy and nine subtypes of legitimacy to respectively legitimize and delegitimize the provision of centralized accommodations to displaced persons.   Frequency (percentage of excerpts coded for delegitimization) Legitimacy type and subtype subtypes of legitimacy to respectively legitimize and delegitimize the provision of centralized accommodations to displaced persons.   Frequency (percentage of excerpts coded for delegitimization) Legitimacy type and subtype In contrast, informants mentioned regulations less frequently when using comprehensibility legitimacy; only 9% of coded comprehensibility legitimacy excerpts were related to regulations. When using comprehensibility legitimacy, informants legitimized actions based on their perception of "what is understandable," based mostly on their past experience. For example, an informant legitimized local government actions by noting that he observed no related problems happening in the past. "I think [the local government] did their very best they could do in this situation. They reacted very fast and in [our state] with most refugees of all states in Germany, with about 200,000, we have no real problems".
Consequential legitimacy was primarily used when informants were assessing the livability of centralized accommodations. Fifty-seven percent (57%) of coded consequential legitimacy excerpts were related to livability. This accounted for 71% of reasons explicitly mentioned by informants when using this legitimacy type. When using consequential legitimacy, informants thought that "the right thing to do" was to provide centralized accommodations with good living standards to displaced persons and focused on benefits provided to displaced persons. For instance, one informant justified her involvement by describing emergency centralized accommodations that her agency wanted to replace, and said, "For the refugees, it is horror. You have zero privacy; they are completely mixed. So we wanted [ . . . ] to let the people move into the [modular buildings]".
To delegitimize the project of providing centralized accommodations to displaced persons, the most frequent types of legitimacy used by informants were comprehensibility, influence, and exchange, as shown in Figure 4. Taken together, these made up 60% of the excerpts delegitimizing the project.
Comprehensibility legitimacy is the most frequent legitimacy type used to delegitimize the project (Figure 4). Based on their experience, 10 informants emphasized that some actions were impossible to undertake (e.g., renting accommodations in a city where there is a severe housing shortage), and 10 informants explained that some actions were better not taken (e.g., taking cultural differences into account when designing facilities). One informant, for instance, delegitimized the construction of new accommodations by saying, "No, no, no, we don't have time".
To delegitimize the project of providing accommodations to displaced persons, the second most frequent legitimacy type used was influence legitimacy (see Figure 4). Informants primarily used this when expressing concern about disadvantages associated with their city, specific neighborhoods, Germany, or different communities. Informants focused, for example, on the fact that providing centralized accommodations to displaced persons is in some cases too costly, challenging, or disturbing for the neighborhood.
In delegitimizing the project, the third most frequently used legitimacy type was exchange legitimacy (see Figure 4). Informants primarily used this one to justify that they were personally not involved in some steps of the project. Informants primarily justified their lack of involvement based on regulations and responsibilities set by their employment contract, manager, and so forth. A majority (60%) of coded excerpts delegitimizing the project while using exchange legitimacy were related to regulations. For example, one informant justified the fact that her agency abandoned a new accommodation project by referring to regulations. "The law says [endangered species] have to be protected. It says that if you build in the outskirts, you are interfering with nature and the landscape". Table 3 summarizes the characteristics of the different types of centralized accommodations for displaced persons in Germany discussed by informants. The table includes informants' perspectives about accommodation types, how informants (de)legitimized the project of providing each type, select justifications offered by informants, and the frequency with which informants described the accommodation types as long and short-term accommodations. To ensure consistency, clear definitions for short and long-term accommodations are used. Excerpts where informants were assuming that displaced persons could live for an indefinite period of time in the discussed centralized accommodations were coded for long-term. Excerpts where informants were assuming that displaced persons could not live for an indefinite period of time in were coded for short-term. Based on the results shown in Table 3, the eight centralized accommodation types categorized in this study can be further classified into five groups:

1.
Sport halls, which have a high percentage of excerpts coded for delegitimization compared to other types.

2.
Former airports and light-frame structures, which were primarily legitimized with consequential legitimacy, have a low percentage of excerpts coded for delegitimization, and they were primarily described as short-term solutions.

3.
Buildings with no major renovations (excluding sport halls and airports) and container housing, which were primarily legitimized with comprehensibility and procedural legitimacy and had an intermediate percentage of excerpts coded for delegitimization.

4.
Modular housing and buildings with major renovations, which were primarily legitimized with consequential legitimacy, had a low frequency of excerpts coded for delegitimization, described as both short and long-term solutions.

5.
Private apartments within centralized accommodations that were primarily legitimized with procedural and consequential legitimacy, had a low percentage of excerpts coded for delegitimization, and were primarily described as long-term solutions (see Table 3 for detailed excerpt frequencies).
We provide below, a summary of informants' perspectives about these five identified categories of accommodations.
First, during the Refugee Crisis, sport halls were used as emergency centralized accommodations. No major renovations were undertaken before the arrival of displaced persons. Indeed, officials intended to use them for only a few months prior to returning them to German schools. Large sport fields were used as common rooms where beds were placed. This accommodation type was the most frequently delegitimized (see Table 3). The delegitimization of sport halls was primarily based on two justifications. First, all informants who discussed sport halls perceived their poor livability conditions, and described this accommodation type as a very short-term solution. One informant stated, "A sport hall is not a shelter where you can stay for a long time normally. It is very hard for the refugees there." Second, two informants emphasized that this accommodation type was hindering the capacity of the schools in the city to operate normally, and that further renovations were needed, at the city's expense, after closing those emergency accommodations. One informant stated, "There have been changes or adaptions made now during the last month while the refugee camp was in the hall. Now when one hall is closed, everything has to be rebuilt".
A former airport was also used to serve as emergency centralized accommodation. This airport was a large, empty building that was partly being renovated to house displaced persons. Separately, the light-frame structures used as centralized accommodations were primarily inflatable domes and large tents. The most recurring legitimacy type to legitimize these structures was consequential legitimacy. This result can primarily be explained by the fact that four informants stated those accommodations were short-term solutions needed to prevent displaced persons from being homeless. "[Tents] were absolutely just for the emergency situation; you can only do that when a lot of people come and they should at least have a place where they don't freeze." This result is verified by the fact that these two accommodation types were largely perceived as short-term solutions (see Table 3): despite legitimizing those accommodations, informants explicitly explained that those accommodations were only very short-term solutions. Buildings, such as former schools, office buildings, and factories, were used as emergency accommodations without being renovated (except for minor renovations, such as painting) prior to the arrival of displaced persons. To serve as emergency accommodations or collective accommodations, container accommodations were newly built in different locations of the cities. Buildings with no major renovations and container housing were predominantly legitimized with procedural legitimacy and comprehensibility legitimacy. The predominant use of procedural legitimacy indicates informants primarily legitimized those two accommodation types by emphasizing that setting up those accommodations corresponded to the right procedure to follow, despite outcomes not necessarily being positive. One informant, for example, supported a city's actions to create new container housing with good living standards, but was not satisfied with the outcome. "I cannot imagine who wants to live there, because they are outside the cities normally, have no connection to the infrastructure. There are nice complexes, good examples done by the city [ . . . ], but I don't think that they will be used after, after these refugees using them." Additionally, the predominant use of comprehensibility legitimacy to legitimize those two types of accommodations demonstrates that informants stated that they understood the government's choice to use those types, based on their past experience but without asserting the associated benefits. For example, one informant said that, based on her experience, containers were an understandable solution to the housing needs, even though it was not a long-term one: "I think the containers are also not a solution but actually for the situation it's OK".
Modular housing and buildings where major renovations (e.g., construction of kitchens and bathrooms) had been undertaken and were (at the time of interviews) intended to serve as collective accommodations. These two accommodation types were more frequently legitimized than the other accommodation types (see Table 3). These accommodation types were primarily legitimized with consequential and procedural legitimacy. Consequential legitimacy was most frequently used when informants were highlighting the fact that modular housing and buildings with major renovations were the centralized accommodations types that provided the best livability. For example, an informant compared the livability of a building that received major renovations to that of emergency accommodations such as sport halls, saying, "Now we are done with the renovations; those housings are regular now; these are more secure shelters. We have now a room for two persons, not for six persons [laughs]." Procedural legitimacy was most frequently used by informants to legitimize modular housing, highlighting the fact that this type of housing can be built quickly while also serving as long-term accommodations. One informant said: "[Modular housing] worked well for this case, because the challenge was to build a lot of housing in a very, very short time." Procedural legitimacy, on the other hand, was used by six informants to legitimize buildings with major renovations, applauding the process of renovating old buildings rather than building new structures. For example, an informant said that renovating unused buildings is better than "starting addressing from the scratch, at a cost of some money".
One particular type of collective accommodation is private apartments for displaced persons in centralized accommodations (e.g., modular housing and container housing). Private apartments were the only centralized accommodation type primarily described by informants as long-term solutions (see Table 3). Informants mostly legitimized private apartments with consequential and procedural legitimacy, suggesting that informants considered private apartments to be a way of providing adequate livability to displaced persons, which was the right procedure to follow. Six informants also used influence legitimacy, stating that the most beneficial centralized accommodation option was providing private apartments to displaced persons for the following reasons: (1) those apartments could be later used by German people, and (2) this accommodation type was a good way to enhance the integration of displaced persons.

Strategies for Informed and Satisfactory Decision-Making
The fact that the project of providing centralized accommodations to displaced persons was more often legitimized than delegitimized could seem counterintuitive given the controversy raised by the Refugee Crisis in Europe in 2015 and 2016 and discussed in Section 1.1. One could expect stakeholders to more frequently delegitimize rather than legitimize the provision of centralized accommodations. The results of this study, thus, indicate an unexpected willingness demonstrated by stakeholders to accommodate displaced persons in centralized accommodations during interviews.
In the context of the Refugee Crisis, the results of this study suggest ways to help stakeholders make decisions during such an emergency situation. First, the results presented in Section 3.1.1. highlight the importance of providing to decision-makers, information about living conditions in accommodations provided to displaced persons (to help them assess accommodations' livability). Such information could be shared via reports and guidelines by the government, nonprofits, and other involved organizations about living conditions in such accommodations. The fact that the livability of accommodations was frequently mentioned by informants using consequential legitimacy (see Section 3.1.2.) indicates that, in order for decision-makers to perceive their decisions as informed, such reports and guidelines should specifically focus on the displaced persons' well-being in such accommodations using local norms of livability (e.g., what is commonly considered a healthy built environment).
The results also highlight that decision-makers, even though they had to react to the emergency situation using their individual judgements of right and wrong, highly relied on regulations. Namely, the fact that a high number of coded procedural legitimacy excerpts were related to regulations suggests that informants primarily assessed the legal framework pertaining to accommodations for displaced persons to decide what "the right thing to do" was. Overall, this shows that regulations provide guidance about the right procedures to follow, even when responses to crises, such as the one studied herein, seem chaotic. We, thus, advise local and state government agencies to not overlook the impact of regulations in such a context of emergency, and to adapt them as dynamically as possible to provide guidance to decision-makers. Finally, the results show the importance of culturally-constructed notions of integration of displaced persons into hosting communities during the making of decisions studied here. To help with decision-making, stakeholders would benefit from information about measures that were successful to the integration of displaced persons in the past. Such information could also help raise stakeholders' acceptance of decisions made by gaining comprehensibility legitimacy, which is critical when making decisions to accommodate displaced persons in centralized housing, as indicated by the results in Section 3.1.2.
Providing to stakeholders information about integration could also help face challenges presented by stakeholders' resistance to the project of accommodating displaced persons in centralized accommodations. Indeed, doing so could help overcome the frequent use of comprehensibility to delegitimize the project. Similarly, and more broadly, communicating about the past successes rather than challenges when accommodating displaced persons (e.g., through press releases) could help with stakeholders' acceptance of the project (by also overcoming delegitimization with comprehensibility legitimacy). Additionally, challenges presented by the frequent use of influence and exchange legitimacies to delegitimize the project can be counteracted by communication strategies emphasizing that accommodating displaced persons in centralized accommodations can have benefits for hosting communities and individual stakeholders.

Choosing Adequate Types of Centralized Accommodation
The results of this study presented in Section 3.2. enable the identification of accommodation types that can be accepted by stakeholders, depending on the associated timeframe. On the one hand, sport halls, former airports, and light-frame structures were perceived as an acceptable option for very short-term accommodation. However, informants were not deeply convinced of their long-term benefits for German cities. Decisions to use such accommodations can, thus, be accepted by stakeholders for short-term use only (typically no longer than a few months in this study). On the contrary, apartments within centralized accommodations were considered a long-term solution that could be beneficial for both displaced persons and German cities. Such accommodations could thus be chosen to increase stakeholders' acceptance of the project of providing centralized accommodations to displaced persons. Intermediately, and for an emergency response to the influx of displaced persons, but for a provision of a longer-term solution than sport halls or light-frame structures, modular housing and renovated buildings were perceived as an adequate solution that can be built quickly. Choosing these specific types of accommodations in the context of an emergency, such as was studied here, can enhance stakeholders' acceptance of such a decision.
This study also indicates that container housing and buildings with no major renovations were perceived by informants as legitimate attempts to provide adequate accommodations (longer-term than sport halls, former airports, and light-frame structures) to displaced persons. However, informants were not convinced about the success of those attempts. In fact, informants' mixed appreciations of living standards in these accommodation types (see Table 3) provided a good indicator that informants had trouble evaluating the effects of the provision of container housing and buildings with no major renovations. Another indicator was that they legitimized related procedures (using procedural legitimacy) rather than their outcomes (rarely legitimized with consequential legitimacy). These results thus show that choosing modular housing or renovated buildings over container housing and buildings with no major renovations could enhance stakeholder acceptance of decisions made to provide centralized accommodations to displaced persons.

Limitations and Future Research
Limitations to this study include the choice of locations and informants, and the investigations' timeframe. All investigations were performed in Germany. The results of this study can thus provide indications about developed countries' institutional responses to sudden international population influxes. Those indications, though, may not be applicable to all developed countries. Indeed, owing to cultural differences, as Hofstede showed, countries vary greatly in their institutional responses [68]. Informants in this study had different perspectives based on various backgrounds [68] (e.g., different types of organizations and responsibilities). Those perspectives were combined to obtain results, and this analysis does not present comparative information about how specific types of institutions (e.g., nonprofits and companies) reacted. A final limitation to the applicability of this study's results may be its timeframe. Interviews were conducted during the summer of 2016. This was at the end of a high influx of displaced persons, after several controversial events that were linked to displaced persons, and prior to state elections by a few months. These circumstances might have affected institutional responses to the influx of the population studied.
Acknowledged limitations of this study also include the small sample size and the use of convenience sampling, which do not allow to draw generalizations across all types of stakeholders (e.g., stakeholders from different cities, with different roles). However, given the limited existing information about decision-making for centralized accommodations in emergency contexts, this study is a valuable contribution to the body of knowledge [69]. For instance, existing studies have validated that meaningful information can be drawn from small sample sets (e.g., [48][49][50]). Many patterns of decision-making and themes discussed in interviews were common regardless of the informants' backgrounds, but future research could assess how such backgrounds affect decision-making in a context such as the one studied here. Future research could also focus on displaced persons' perspectives, as they might significantly differ from those of the informants discussed here, and because they were the only key stakeholders identified in introduction who were not included in this study.

Conclusions
In this study, the authors understand the provision of centralized accommodations to displaced persons to be an engineering and construction project. The way key stakeholders perceive the legitimacy of this undertaking is important to project success. For example, the perspectives of individuals within organizations can affect the efficiency of social collaborations (such as [43]). Still, accommodation decisions for large numbers of displaced persons may be controversial, meaning gaining and maintaining legitimacy can be arduous. As such, this study contributes to filling this gap by assessing perceptions of the legitimacy of the technical project of providing centralized housing among those stakeholders who were involved in this work in Germany between 2015 and 2016. In addition, this study shows that the qualitative approach used here is useful for researchers studying institutional responses to controversial or sudden changes (e.g., regulation changes, sudden migration, etc.).
Interview data were analyzed using qualitative methods to explore how informants used different types of legitimacy, and how those types of legitimacy were linked to topics such as the livability of accommodations or the different types of accommodations. The most frequently mentioned reason for legitimizing the project of providing centralized accommodations to displaced persons was that of achieving an adequate standard of livability. In contrast, the most frequently mentioned reason to legitimize decisions made to not accommodate displaced persons were regulations. The legitimacy types used by informants to legitimize the project of providing centralized accommodations for displaced persons were primarily moral (mostly procedural and consequential), and those used to delegitimize the project were primarily cultural-cognitive (mostly comprehensibility). In other words, different forces legitimized and delegitimized the centralized accommodation project. The legitimization of the project was mostly based on individual convictions of right and wrong (moral legitimacy), while the delegitimization of the project was mainly based on an understanding of the fact that, in some cases, it is impossible to provide centralized accommodations to displaced persons (cultural-cognitive legitimacy). For example, some respondents legitimized the housing project by noting that it was morally wrong to leave displaced persons homeless, while some delegitimized the project by noting that it was simply impossible to rent such a large number of apartments in the context of extremely low vacancies.
The results of this study indicate that the different centralized accommodation types were not legitimized in the same ways. Sport halls had the highest percentage of excerpts coded for delegitimization, and were perceived as providing low livability (e.g., no privacy). In contrast, modular housing and renovated buildings had the lowest percentage of excerpts coded for delegitimization, and were perceived as providing high living standards and long-term accommodation. Respondents mainly saw light-frame structures and former airports as legitimate when used as short-term solutions that prevented displaced persons from being homeless. In contrast, container housing and buildings with no major renovations were seen to be a legitimate option for longer periods of time, as this was better than "doing nothing"; compared to other options, however, the informants were unsatisfied with the results.
Arising from these results are recommendations to help with the making of informed decisions that can be accepted by the stakeholders: (1) Provide to stakeholders, information (e.g., through reports and guidelines) about accommodations' livability with a focus on displaced persons' well-being to help them use consequential legitimacy.
(2) Adapt regulations to the emergency situation when possible because they are critical in stakeholders' perceptions of right and wrong (i.e., when using procedural legitimacy).
(3) Provide information about past successes in accommodating displaced persons in centralized housing (to ease the use of comprehensibility legitimacy), with a specific focus on integration.
(4) Carefully choose types of centralized accommodation based on the intended associated timeframe-short versus long-term. Depending on the timeframe, stakeholders legitimized different types of accommodations.
(5) For long-term accommodation, prefer fully renovated buildings or modular housing (associated with consequential legitimacy) to container housing or buildings with no major renovations (associated with procedural legitimacy).
These recommendations are expected to be applicable to European high-income countries hosting internationally displaced persons, but also more broadly to high-income countries worldwide facing sudden influxes of displaced persons. For instance, the results of this study could be applicable to U.S. cities responsible for hosting sudden nationally displaced populations resulting from disasters-such as the 2018 displacement in California resulting from the Camp Fire [70].