Outsourcing and backsourcing in a time of crisis – experiences of double opportunism

ABSTRACT

This article addresses the question of how public organizations with limited capacity manage to handle large scale crises. The case analyzed in this article is the sourcing of accommodation during the migration crisis in Sweden of 2015–2016. The event was a game-changer in Swedish migration policy following a more hesitant policy than before, even though new and substantial migration crisis have followed since then. Results show how the legal framework for procurements was not adapted for extraordinary events. The Swedish state was forced to pay exaggerated prices as private contractors took advantage of the acute demand. The study develops our theoretical understanding of the sourcing process during crises and show how the short-term perspective and uncertainties that characterize crises encourage both suppliers and procurers of services to act opportunistically.

KEYWORDS:

• Backsourcing
• outsourcing
• double opportunism
• New Public Management
• agency theory
• migration

1. Introduction

In recent years countries in Europe have witnessed several major crises and large dramatic events such as the financial meltdown of 2008–2009, the migration crisis of 2015–2016, the COVID pandemic of 2019–2021, and the ongoing war in Ukraine since 2022 (Christensen & Lægreid, 2022; Niemann & Speyer, 2018; Niemann & Zaun, 2018; OECD, 2020; Simsa, 2017). All these extraordinary events have called for large-scale public interventions and governmental crisis management (Bouckaert et al., 2020; Genschel & Jachtenfuchs, 2018).

With the need for substantive public intervention increasingly evident, the capacity of public organizations to handle extreme events may simultaneously have been impeded by New Public Management (NPM) reforms such as deregulation, marketization and outsourcing of public services to private sector suppliers (Funck & Karlsson, 2020; Lapsley, 2009). The implementation of such reforms has made the public sector increasingly dependent and reliant on resources and support from private suppliers, not at least to cope with crises. Unfortunately, the implications of NPM reforms for crisis management in the public sector have hitherto received scant attention in the literature (Christensen & Laegreid, 2011; Christensen & Lægreid, 2022; Christensen et al., 2015). In a world where public authorities are increasingly reliant on sourcing the management of their everyday services (Funck & Karlsson, 2020), we know far too little about the challenges of sourcing in life and death situations of great importance where there is limited time for planning and issuing contracts. It is therefore important to find out more about how crises are handled under market-like conditions, primarily in order to be able to parry the difficulties in a better way in the future.

Even though many countries rely heavily on sourcing to manage extraordinary and unforeseen events, research on how sourcing is used by public organizations to handle crises of this magnitude is to our knowledge lacking (Christensen & Laegreid, 2011; Christensen et al., 2015; Kastanioti et al., 2013; Nikolaou & Loizou, 2015). The aim of this study is to contribute to the existing research on sourcing by investigating specific challenges faced by public outsourcing and backsourcing in times of crisis and how crisis influence opportunism among both principals and agents from an Agency theory perspective.

The study focuses on a case where an externally generated crisis unexpectedly tested the abilities of a small country to its limits. The case is important because it faced heavy criticism and marked a pivotal shift in Swedish migrant policy going from openness and hospitality to a more restrictive policy (Hagelund, 2020; Krzyżanowski, 2017). From a European perspective, Sweden has had a generous approach toward asylum seekers in recent decades (SOU, 2017, p. 12). During the period 2000–2012, Sweden, a country of 10 million inhabitants, received around 30,000 asylum seekers per year. However, there was an unexpected and dramatic increase in the numbers in 2013–2014, reaching 163,000 in 2015, more than half of them during a period of just two months (SOU, 2017, p. 12). As the Swedish Migration Agency, the state authority responsible for providing accommodation for newly arrived asylum seekers, had nowhere near the housing capacity needed for this number of people, it hastily resolved the situation by sourcing a large volume of accommodation services from the private market (Hagelund, 2020; RiR, 2016; SOU, 2017, p. 12).

Understanding this is crucial as countries require well-established routines and structures to increase the ability to deal with emerging humanitarian crises in a short time, and the development and refinement of theoretical tools are necessary to enhance our understanding and analyze the meaning of contradictions and paradoxes when dealing with this type of event. Overall, this study aims to enhance the ability of countries to manage similar scenarios in the future.

The remainder of the paper is organized as follows. Section 2 presents agency theory and provides a theoretical perspective on sourcing. Section 3 provides a setting and an empirical background on the evolution of the migration crisis from a European and Swedish perspective. Section 4 presents the method applied along with the participating respondents. The results are presented in section 5 and discussed in section 6. Finally, the study’s conclusions are presented in section 7, along with suggestions for future research.

2. Theoretical framework

In this section, we will present the two theories applied, Agency theory and Transaction Cost Economic Efficiency theory (TCE), to analyze the sourcing decisions of the Swedish state and the changing relationship between the Swedish state, the Swedish Migration Agency (SMA), and its suppliers of accommodations during the 2015–2016 migration crisis. The theories aid in understanding and describing the emergent situation, providing a framework for its analysis and explanation.

2.1. Agency and TCE theory on opportunism

We will use agency theory, one of the most commonly used theories in sourcing management studies, to explore the relationships between procurer and suppliers in times of turmoil and crises.

Agency theory is frequently used to understand the relationship between a procurer of services and its suppliers. The theory makes two important assumptions. First, it assumes a conflict of interests (conflict level) between the purchaser of services (the principal) and the suppliers (the agents) (Baiman, 1990). Second, that there is an information asymmetry between the two actors, which entails the agents having access to more information than their principals and that the former tend to take advantage of their informational advantage (cost and control) by acting opportunistically in the sense of taking advantage of opportunities to the benefit of oneself with little regard for principles or consequences for others (Williamson, 1979). Furthermore, agency theory highlights two problems associated with information asymmetry (Lane, 2005). The first is that the principal finds it difficult to observe and monitor all the agent’s hidden actions. This means that the agent acquires a gradual information advantage over its principal. To prevent this, the principal needs to invest resources in information gathering, knowledge acquisition and control. However, checking all the agent’s actions can be laborious and costly for the principal. The second problem rises when the outcomes of the agent’s actions are influenced not only by their efforts but also by how others use this hidden information. If the agent blames others, it is difficult for the principal to know whether the result is due to the agent’s contribution or if it is due to the efforts of others (Lane, 2005). If the principal suspects that both problems may arise with an agent, it is difficult to base the remuneration solely on the agent’s contribution, as other factors might also need to be considered (Petersen, 1993). It follows from this that sourcing always comprises an asymmetric relationship between the principal and the agent that is influenced by the trust and confidence that exists between the two actors (Baiman, 1990). If the principal loses confidence in the agent or if the principal finds the costs higher or the service quality lower than expected, it can decide to terminate the contract and bring the activity back in-house called backsourcing by Young and Macinati (2012).

A conflict of goals is thus Inherent in the principal-agent relationship. The conflict occurs as the actors are perceived as profit maximizers that try to maximize their own profit even at the expense of others (Foss & Stea, 2014). In addition to these conflicting goals, there is the issue of information asymmetry that enables the agent to act opportunistically at the principal’s expense (Williamson, 1979).

Opportunistic behavior by suppliers is one of the most common risks identified in outsourcing decisions (Fleta-Asin et al., 2019; Lohmann & Rötzel, 2014), and in some cases it causes principals to not outsource (Greenberg et al., 2008).

Three types of opportunistic behavior have been described in literature (Connelly et al., 2011): adverse selection, moral hazard and holdup problems. Adverse selection occurs prior to the contracting phase, when agents misrepresent the quality and/or quantity of their services in their efforts to win a contract. Moral hazard is when the agents do not fulfill contracts and try to get away with it because the information asymmetry makes it difficult for principals to monitor fulfillment of the agents’ contracts. Hold-up problems occur when agents change the terms of contracts, such as increasing prices, due to the principals’ dependence on their services, by increasing prices for example, to be more favorable.

Agency theory has mainly been developed for and applied to orderly conditions characterized by stability, planned procurements and long-term contracts (Eisenhardt, 1989; Petersen, 1993). Conditions are radically different during crises, characterized by uncertainty, unpredictability and the need for urgent measures with little opportunity for procurers to plan procurements and establish long-term relationship with their suppliers. We will try to determine the extent to which agency theory is applicable in explaining the relationship between principal and agents in this type of irregular and unstable situation. Previous research has discussed principals’ opportunistic behavior of from a theoretical perspective, in for example transformational outsourcing relationship (Rebernik & Bradač, 2006) and on theoretical grounds (Wagner, 2019). From an Agency theory perspective Wagner (2019) for example argues that the opportunism of principals and agents differs fundamentally. Whilst agent opportunism rest on information asymmetry, principal opportunism depends on power asymmetries in favor of the principal (Wagner, 2019). To our knowledge, empirical research covering both agent and principal opportunism of in sourcing, so called double opportunism, is scarce (Porcher, 2021; Steinle et al., 2020). Previous research has primarily focused on opportunism by suppliers or, less frequently, on detrimental actions by suppliers or, although less frequently, as detrimental action by procurers (Steinle et al., 2020). Eisenhardt (1989, p. 71) emphasizes that agent-principal theory is particularly applicable in relation to opportunism, contractual problems, conflicting goals and low measurability (market situation). Eisenhardt (1989) also emphasizes that the theory needs to be ‘expanded to a richer and more complex range of contexts’ (p. 71). For example, is agency theory relevant to understanding how the principal and agent manage sourcing under the unpredictability that occurs during extraordinary circumstances.

Complementing agency theory, TCE theory provides us with reasons for why opportunistic behavior might arise in the relationship between a principal and an agent. TCE theory suggests that misaligned incentives promote opportunistic behavior as, typically, every dollar out of the principal’s pocket goes into the agent’s pocket (Lacity et al., 2011; Williamson, 1979). In turn, principals are incentivized to demand more services from agents without wanting to pay more, and suppliers are incentivized to squeeze as much profit from existing contracts or to sell additional services to increase revenues. Opportunism should be at its highest when the principal cannot specify or does not know what it wants and cannot accurately assess whether suppliers are honoring their commitments (Ellram et al., 2008; Williamson, 1979). In addition to the inherent conflict of interest and misaligned incentives between the actors, TCE theory suggests that asset specificity, uncertainty and frequency of exchange give rise to opportunism (Williamson, 1979). Asset specificity concerns the specific investments made by the parties in the relationship and the uniqueness of the services that the agent provides. Frequency of exchange refers to how often the procurer initiates a transaction with the agent while uncertainty is the volatility, unpredictability and ambiguity that is associated with the outsourcing (Williams, 1979).

This study contributes to the existent research by presenting empirical findings on how the uncertainties that characterize crises may influence opportunistic behavior of both the agents and the principals. By examining the applicability of agency theory in the case of sourcing accommodation during the Swedish migration crisis of 2015–2016, the study explores whether agency theory can be modified and developed to advance understanding of how sourcing in the public sector is managed under extraordinary circumstances.

3. Background: the migration situation in Europe and Sweden 2015–2016

International research shows that many European governments were profoundly unprepared to handle the influx of refugees during the 2015–2016 migration crisis (Genschel & Jachtenfuchs, 2018; Niemann & Speyer, 2018; Niemann & Zaun, 2018; Simsa, 2017). Taking in refugees and asylum seekers and finding them proper accommodation is challenging under normal circumstances, but when the number of refugees seeking to enter Europe exceeds one million in just a few months (market situation), new and innovative measures are crucial for states stripped of their resources in the aftermath of privatization and New Public Management (NPM) (Myrberg & Lægreid, 2019). European welfare states such as Greece, Austria, Germany and Sweden, which received the lion’s share of the refugees trying to enter Europe, were forced to rely not just on public resources but also civil society and private corporations (Myrberg & Lægreid, 2019; Simsa, 2017). The 2015 migration crisis has been followed by the Ukrainian refugee crisis and new directives have been presented in the EU, i.e the Temporary Protection Directive (TPD), allowing Ukrainian citizens the right to work, live, access healthcare, education and housing in the EU, without the requirement to go through lengthy asylum procedures (Bosse, 2022). In contrast, the 2015 refugee crisis rapidly escalated, in the absence of any visible government support, Greece became entirely dependent on volunteers, whilst aid agencies warned of the severe humanitarian conditions for thousands of stranded refugees (Smith, 2016). In Austria, a country with a similar type of ambitious welfare system and population size as Sweden, the massive influx of asylum seekers forced the government to rely on NGOs and volunteers as a ‘gap filler’ when governmental resources were insufficient to provide shelter and accommodation (Simsa, 2017). Germany also relied on voluntary engagement to help those who arrived (Tränhardt, 2018), as finding accommodation for the large numbers of asylum seekers traveling via Hungary, Austria and the Balkans posed major challenges for state and local authorities (Tränhardt, 2018). In Italy, which received the bulk of the refugees from North Africa via the Mediterranean, the reception and housing of refugees were partly outsourced to private contractors and accommodation became a lucrative business, with landlords and hotel managers turning their available space into housing (Savino, 2016). As in Italy, housing services in Sweden were acquired in a similar fashion, relying heavily on sourcing from the private sector. The public procurement in Sweden and Italy both resulted in scandals arising from corruption and the influence of organized crime (Myrberg & Lægreid, 2019; Savino, 2016). Mostly the accommodations provided for the 163.000 refugees (of which 35.000 were unaccompanied children) were either rented apartments with self-catering, corridor accommodation that the Migration Agency rented directly from property owners that was self-catering or temporary accommodation procured by the Migration Agency, usually with full pension (Migration Agency’s annual report for 2015; RiR, 2016). It was mainly the latter type of accommodations that were terminated with the decreased inflow of refugees after enforced stricter border controls by the Swedish government in November 2015. Just a few months after the 2015 crisis the Swedish government also proposed a refugee policy adjusted to the EU minimum level.

To conclude, the European countries were all unprepared and lacked sufficient resources in the initial phase of the crisis, and they handled the situation differently: some such as Austria by relying mainly on public resources and volunteers while others such as Sweden who relied more on sourcing resources in the private market.

4. Methodology

4.1. The case

This case study focuses on how procurement was used as a strategy to manage the Swedish migration crisis of 2015–2016 (Carson et al., 2001; Yin, 2003). Sweden is an interesting case as it was the Country in the EU that received the most refugees during the migration crisis in proportion to its population (Hagelund, 2020; RiR, 2016; SOU, 2017, p. 12). Additionaly, Sweden is also one of the countries in Europe that have most fundamentally and rapidly privatized public services and implemented a new public management regime. Moreover, Sweden is a particularly interesting case as it has a policy for crisis management that is based as far as possible on the fact that procedures used under normal conditions should also be used to manage crises (Becker & Bynander, 2017). Sweden also shares legislation pertaining to public procurement with other EU countries.

4.2. Data collection

Data were collected from interviews with civil servants from the immigration authorities in Sweden. Data were gathered shortly after the end of the crisis in 2018–2019, recent enough for all key actors to easily recollect events but still enabling us to obtain a good overview of the whole process (Kvale & Brinkman, 2009). Interviews and documents have enabled us to study actors and to reconstruct the process. The reconstruction has enabled us to divide the process into different phases, from the initial decision, via the acute management and a stabilizing situation and onward to the operation’s conclusion (Kvale & Brinkman, 2009).

The interviewees included senior officials and administrators at the Government Offices; senior officials, administrators at the Migration Agency; company executives for the companies that provided accommodation units; and refugees who had experience of these types of housing. These interviews were held both face to face and over the phone (Cachia & Millward, 2011). With the consent of the interviewees, the conversations were recorded. The conversations were transcribed verbatim (353 pages) and saved in anonymized form. A total of 24 interviews were conducted (14 men and 10 women), lasting on average for 25–45 minutes (771 min in total). The participants’ written consent was obtained in all cases. The interviews were semi-structured and followed an interview guide (see Appendix A with question areas (see Appendix B for detailed information concerning the interview persons). The purpose of collecting data from a range of actors was to enable a more detailed description of the development and the experiences that the actors had during the course of the process (Kvale & Brinkman, 2009). About half of the interviewees were managers or in other senior functions. The mix of interviewees meant that we obtained a wide and varied picture of the process. The asylum seekers were recruited with the assistance from the Swedish Red Cross contact center for refugees.

In the interviews, the senior officials, administrators, business people and migrants all openly shared their experiences of managing the challenges that arose during the turbulent process. In order to verify the interviews, we also inspected relevant documents from the Migration Agency that had been produced in connection with the procurement. The documents included procurement details, contracts, descriptions, policy documents etc. The documents were used to verify how the organizations acted and managed the process. The two sources of data complemented each other.

The interviews were transcribed and coded with the aim to look for patterns, overlaps, and meaningful descriptions that laid the foundations for creating relevant themes with relevance for the aim and research questions (Tracy, 2019). The first step entailed finding meaningful units that formed categories. These comprised key sequences that were perceived to be important in order to understand the sourcing process and the transition of the relationship between the agent and the suppliers (Krippendorff, 2019). Nodes were also created, linking them to our theoretical framework (TCE and the Agency theory). After several read-throughs, the material was analyzed abductively, where we moved from empirical observations to theoretical concepts and back (Krippendorff, 2019). This enabled us to perceive how the sourcing process could be gradually described and explained. In a second analytical step, we were able, as we gradually observed how the different parts were linked, to compile the different parts into a whole. This approach enabled us to describe and explain the behavioral patterns that emerged in relation to sourcing and opportunistic behavior.

5. Results – four phases of the sourcing process

The results are presented in relation to the main phases of the crisis and are structured to focus on the aim of this study, to contribute to the existent research on sourcing by investigating specific challenges faced by public outsourcing and backsourcing in times of crisis and how crisis influence opportunism among principals and agents from an Agency theory perspective. For this purpose we use the 2015 Swedish Migration Crisis. By focusing on this case we will be able later on to discuss lessons to be learned from this example about sourcing in crisis and the theoretical contributions that can be gained to the agency theory.

5.1. The pre-crisis phase

Why did the public authorities decide to manage the migration crisis through outsourcing and procurement? As it turns out, key actors at the Swedish Migration Agency (SMA) did not perceive much of a choice. They describe the increase in the numbers of migrants seeking asylum during 2015–2016 as an unforeseeable and exceptional event, which surprised the SMA and put the whole country in a problematic situation. The SMA did not have the resources and was not prepared to manage a housing challenge of this magnitude (RiR, 2017). The SMA needed to act swiftly to deal with a situation that seemed close to getting out of control. One of the section managers explained: ‘We are set up to cope with normal conditions, all authorities, the whole community, but when something exceptional happens, we are not prepared and can’t afford it either.’ (#4)

As the SMA was unprepared and did not have sufficient capacity (according to the parliaments auditory service (RiR, 2017) and officials at the agency) it was forced to go beyond its organizational boundaries and procure the required accommodation capacity from external suppliers on the private market. ‘Purchasing from external contractors was the only way.’ (#5).

Furthermore, officials at the Department of Justice as well as at the SMA maintained that it would be unrealistic to have accommodation capacity in reserve for what was perceived to be a highly unlikely event. Governmental officials at the Department of Justice stated that the Swedish government could not afford to maintain empty housing of this magnitude in case of extreme events. Relying on external capacity was perceived to be a more realistic and cost-effective approach. According to an official at the Justice Department, the strategy of procuring housing from private suppliers provided the SMA with the flexibility needed to adjust its accommodation capacity at short notice due to a changing inflow of asylum seekers: ‘There are some substantial benefits of contracted accommodation, and it is this flexibility that can be scaled up and dimensioned at short notice’ (#2). Flexibility was frequently mentioned by centrally positioned SMA officials as a significant advantage of sourcing in comparison to in-house accommodation capacity.

5.2. The procurement phase

Public procurements are time-consuming procedures, regulated in detail by the Public Procurement Act (LOU). In urgent situations, such as the 2015–2016 migration crisis, it can be hard to reconcile the time-consuming legal process of public sourcing with the need to act as swiftly as the situation requires.

The respondents at the SMA stressed that there is a clear underestimation at governmental level regarding the complexity involved in procuring accommodation from external providers. As one of the officials put it: ‘I would say that in this kind of crisis situation, where speed is of the essence, … [The Swedish Procurement Act] is not adapted to those conditions.’ (#7)

The procurement decisions were also frequently appealed by private contractors competing for lucrative contracts. Respondents admitted that some of the underbidders will always, rightly or wrongly, perceive that they have been disadvantaged. As the agencies purchased about 60,000 temporary accommodation units from more than 1,000 different private suppliers during the crisis, the number of legal appeals represented a significant administrative burden for the SMA.

Some of these legal disputes not only concerned the process of selecting which supplier would be awarded the contracts, but also issues concerning allocation of responsibility between the purchaser and the supplier. The crisis meant that the principal (SMA) was pressed for time and lacked experience in managing these levels of very large procurements from hundreds of different suppliers. Respondents indicate that the SMA also neglected to check that the terms in the contracts were met by the suppliers. The information asymmetry between the SMA and its suppliers in this phase was clearly advantageous to the suppliers, which in some cases acted opportunistically by not fulfilling their part of the agreement. In some cases, the contracts were also insufficiently specified. For example, one respondent (#8) recalls a disagreement regarding whether the SMA or the private suppliers should provide toilet paper for the asylum seekers. Some officials at the SMA were critical regarding how they conducted these procurements, especially in the early phase of the crisis. However, this improved over time as the agency learned from its experiences.

Not only did the legal framework for public procurements slow down the process, but it was also perceived to deprive officials from taking into account vital concerns other than price and quality, and the quality of these kinds of services is of course already difficult to assess. One example of a neglected factor was the geographic location of the accommodation facilities. This shortcoming resulted in questionable decisions concerning accommodation in locations that were unsuitable for housing such a large number of asylum seekers, especially sparsely inhabited rural municipalities where the social infrastructure such as schools, health care and libraries is not set up for a large increase in the number of inhabitants.

The legal framework of public procurement also prohibited the SMA from communicating with local stakeholders before the official decision date. Municipalities where it was likely that asylum seekers would be accommodated were therefore deprived of the opportunity to prepare for their reception, including organizing preschools, schools, health care, dental care and social services etc. This was very frustrating for many local governmental officials, with one of them stating: ‘The SMA called on Friday and said … 200 people will move to this small village on Monday, 50 of whom are schoolchildren and would like to start school on Monday.’ (#9) The SMA’s interpretation of the legal framework of public procurements meant that it was unable to centrally plan and decide on how best to distribute the accommodation across the country. To summarize the implications of the legal framework, as interpreted by SMA administrators, it is not flexible enough to adapt to a situation where large-scale procurements are needed at short notice.

5.3. The adjustment phase

To cope with the significant increase in refugees during the peak of the crisis in late 2015, the SMA decided to lower the requirements it placed on private accommodation facilities in order to attract more suppliers. Several prerequisites were eliminated, for example, conditions concerning the financial situation of suppliers and subcontractors. At the start of 2016, the SMA had contracted more than 1,000 private suppliers, many of which had no previous experience in providing accommodation for people in socially vulnerable situations or which were not financially stable. Consequently, quite a few of the suppliers where unreliable, and, according to officials at the SMA, the quality of the services was uneven. ‘Some accommodation that we have procured has been very good … that has not been the case for others’. (#5)

Many of the suppliers underestimated the complexity of the services. Some wrongly perceived that the contract would only require them to provide housing and food. However, providing accommodation for refugees proved to be a complex task. Several of the private suppliers and asylum seekers interviewed identified conflicts, insecurity, frustration and violence. Many suppliers were forced to hire security companies and, on some occasions, call the police to ensure safety and order. Several of the private suppliers experienced vandalism.

Given the urgency of the situation, the SMA focused on meeting the acute demand for accommodation. In combination with the SMA’s limited resources available to evaluate the suppliers and inspect the facilities, this priority meant that many contracts were given to suppliers that were unprepared for the assignment. The fact that many suppliers were inexperienced and fundamentally unqualified to provide the services expected of them did not occur to SMA administrators until after the contracts were granted.

According to a report from The Swedish National Audit Office (RiR, 2016), over half of the audited accommodation facilities contracted by SMA in 2015–2016 had major unfulfilled requirements, and in many cases these deficiencies were never resolved (RiR, 2016).

As many of these suppliers were unprepared and unqualified to perform their commission, the SMA was forced to terminate contracts in advance and backsource a large number of accommodation units. Backsourcing of accommodation became increasingly common during 2016, after border controls had been imposed and with the decrease in numbers of asylum seekers. Having less need for accommodation and more time to inspect the facilities, the SMA was more motivated to fulfill its responsibility as a principal and to ensure that public resources were used efficiently. The SMA started to inspect facilities more thoroughly, checking that suppliers were meeting all their contractual demands. Not only was the SMA more inclined to enforce its demands once the influx of refugees had decreased, it also decided to enforce stricter demands on the accommodation facilities that it purchased. This change of approach caused further appeals and legal disputes, marking a new phase in the relationship between the SMA and its suppliers. The balance of power began to tip toward the SMA as demand for supplier services decreased.

In many cases, it was unclear to the suppliers that they had only been granted temporary contracts that could be canceled if they did not fulfill the pre-specified requirements. In the meantime, agreements with suppliers could be signed further down the ranking in the list of bidders that had already met the requirements. In practice, this meant that sometimes the agency did not sign agreements with suppliers according to the ranking published in connection with the award decision, but according to the time that the suppliers fulfilled the requirements. This strategy triggered even more legal disputes.

5.4. The re-integration phase

After implementing stricter border controls in November 2015, the influx of refugees into Sweden decreased from about 40.000 a month to less than 10.000 refugees in December (RiR, 2016). As the inflow of asylum-seekers continuously decreased the following months the SMA decided to successively terminate contracts with their private suppliers during 2016 (RiR, 2016). Asylum seekers were transferred, either to temporary private homes or to the SMA’s own accommodation facilities. Terminating private suppliers’ contracts was a process that had to follow the general legal framework of the Public Procurement Act, which meant that the SMA was forced to terminate contracts according to the initial ranking of suppliers and not according to their own preferences. Consequently, the SMA was unable to prioritize accommodation facilities in strategic locations and terminate poorly located housing.

This led to large numbers of asylum seekers having to be moved around several different accommodation facilities as contracts with suppliers were terminated in the prescribed order. One of them lamented: ‘I have been moved to several locations in a short period. I would have preferred to stay at one place.’ (#23) Many of the refugees interviewed had been forced to move 3–4 times, and this could be especially frustrating for families as the children would have to change schools and separate from newly found friends. ‘We were frustrated as we were not informed in advanced that we were expected to move to another location,’ one asylum seeker explains. (#20)

When discussing the termination of accommodation facilities and the transfer of refugees to new facilities, it was evident that SMA officials, private contractors and asylum seekers had different views on how the process was planned and carried out. The suppliers often expressed more critical views, especially regarding the decision to terminate contracts and the backsourcing of accommodation capacity: ‘It has been catastrophic. They (SMA) have acted as they wanted.’ (#13). According to several of the private suppliers, these decisions were announced at short notice ’… some of the applicants have become very upset and tired of moving. I think that has been the big negative for the applicants.’ (#14). Some SMA officials also acknowledged that the relocation of asylum seekers to new accommodation may have affected them badly.

6. Discussion

6.1. Lessons learned

The European migration crisis of 2015–2016, which impacted Sweden more severely than most Western European countries, is an interesting example of what can happen when a government is reliant on sourcing during a crisis (Hagelund, 2020). The challenges that Sweden faced in finding accommodation for asylum seekers during this crisis are also theoretically interesting as they provide examples of the advantages and shortcomings of sourcing during a crisis (Carlsson-Wall et al., 2020). The case analyzed in this paper also highlights the relevance (but also shortcomings) of agency theory as an analytical tool to explain the events (Baiman, 1990; Lane, 2005; Petersen, 1993).

By relying on sourcing, the Swedish state was able to adjust and cope with a situation for which it was utterly unprepared (SOU, 2017, p. 12). Temporarily depending on private accommodation suppliers allowed the government to avoid substantial costs associated with maintaining a permanent, yet unused, housing capacity (RIR, 2016, 2017). However, the results also show that there is a high price attached to dependence on private suppliers in a crisis. The Swedish state – i.e. Swedish taxpayers – was forced to pay exaggerated prices for services provided, as private contractors acted opportunistically, taking advantage of the substantial demand for accommodation by increasing prices. On the other hand, not keeping housing capacity in reserve may also be considered as opportunistic and should be included in the after-math analysis of costs. Another side of the coin was the dependence on external suppliers deprived the Swedish government of operational control and the ability to ensure an even quality of the services provided during the peak of the crisis.

The fact that a public procurer needs to consider the legal framework that restricts public sourcing may, in this case, have caused a poor adaption to what was required for swift and effective action. The study illustrates the fact that many procurement decisions were subject to appeals and delays. The huge procurements also impacted on the SMA’s ability to monitor the quality of the accommodation and check that it met the pre-set demands, resulting in contracts that were terminated in advance, thus causing even more appeals and more time-consuming legal disputes. Another source of legal appeals arose when the SMA sought to reduce the number of private accommodation units as the number of refugees seeking asylum decreased (RIR, 2017). On some occasions, the order in which these contracts were terminated deviated from the order prescribed by the legal framework on humanitarian grounds, causing further appeals (RIR, 2017). Following the prescribed order of termination meant that the asylum seekers had to be moved several times. The legal framework regulating procurements was clearly perceived as inadequate, to deal with crisis situations, by the SMA. The study further shows that the initial information asymmetry that initially arises during contracting is a decisive factor. The negative effects could be reduced by having stable companies contracted in advance, ensuring their readiness for crises. Additionally, authorities should enhance their preparedness for managing certain predictable crisis scenarios. It is crucial for authorities to boost their capacity to continuously monitor that the contractors fulfill and adhere to the agreements entered.

6.2. The four phases

The main theme of this study is the sourcing process in times of crises. In previous research, sourcing has mainly been described as a planned and relatively stable process (Damanpour et al., 2020; Johansson, 2015; Veltri & Sunders, 2008). This study reveals that such a simplistic description is not universally applicable, particularly not to sourcing in dramatically evolving crisis situations (Carlsson-Wall et al., 2020). The case study revealed an acrimonious process, evolving through four distinct phases (see Table 1). In this article, the phases are distinguished by the changes in the relationship between the SMA (the principal) and its external private suppliers (the agents). The development of relationships throughout these phases can be further analyzed by applying concepts frequently used as ‘Power balance,’ ‘Costs and control,’ ‘Conflict level,’ and ‘Market situation’ (Eisenhardt, 1989; Hood, 1995; Lapsley, 2009; Widman, 2016).

Table 1. The sourcing phases.

	(1) The pre-crisis phase	The crisis occurs	(2) The procurement phase	(3) The adjustment phase	(4) The re-integration phase
(A) Major focus	The make-or-buy decision		Rapidly acquire resources	Consolidate with suppliers’ terms	De-sizing and re-integration
(B) Power balance	Status quo		Tips in advantage of the supplier	Tips in the advantage of the procurer	To the advantage of the procurer
(C) Costs and control	Low costs, high level of control		High costs, low level of control	High costs, increasing control	Reduce costs and increasing control
(D) Conflict level	Low		Increasing	High	High
(E) Market situation	Functioning		Strained	Strained	Resources exceed the need

In the first phase, the pre-crisis phase, the main focus concerned the extent to which the public principal should build up internal resources to manage a crisis or rely on sourcing from external suppliers. In this situation, the make-or-buy-decision was driven more by economic efficiency than by political considerations. The principal (SMA) mainly relied on internal accommodation resources, but also had a limited number of long-term agents (private accommodation suppliers) with whom it had established long lasting relationships built on confidence and trust. However, only a small proportion of the accommodation services were procured before the crisis. In this phase, relations between the principal (SMA) and the agents could be depicted as symmetrical and stable and there was thus a low level of conflict between the agent and the principal. When needed, the principal was easily able to acquire accommodation services at low cost and in an orderly fashion among several bidders in the private market. The relationship between the principal and the agent thus functioned according to theoretical (agency theory) expectation (Baiman, 1990).

In the second phase, the procurement phase, the crisis had emerged and the relationship between principal and agents was shaped by the crisis and the increasing demand for accommodation services. The balance of power during this phase tips in favor of the agents, as the principal is forced to rely on external providers to be able to fulfill its governmental responsibilities. The cost of acquiring these services rises as the demand for accommodation services begins to exceed the resources available in the regular private market. The level of conflict between the principal and the agents is low in this phase but increasing as disappointed bidders appeal against some of the public procurements. Lack of time made it difficult to find suitable companies with previous experience in refugee reception. Our study also reveals, contrary to theoretical expectations, that the procuring authority (principal) needs to act opportunistically to resolve the situation that has arisen.

In the third phase, the adjustment phase, the balance of power becomes more equal between principal and agents as the crisis begins to stabilize and there is an increased availability of external resources. This gives the principal the opportunity to be more selective in terms of which agents to procure from, and to be stricter in following-up and enforcing the terms in the contracts that have been established. The level of conflict between the agents and the principal further increases in this phase. Legal disputes increasingly arise as contracts are terminated when agents are shown to be unable to fulfill their contractual responsibilities.

In the fourth and final phase, the re-integration phase, the crisis begins to de-escalate, with demand for private external service providers rapidly decreasing. The principal is focused on reducing costs, terminating contracts with external suppliers, and re-integrating services (backsourcing). The balance of power is clearly in favor of the public procurer, which increases its effort to wind-up the services procured as soon as possible, while the external service providers are interested in keeping their contracts in order to be able to pay off their investments. In this phase, the level of conflict peaks, generating large volumes of time-consuming legal disputes overcompensation for damages and premature contract termination. The termination of contracts and re-integration of services also impacts on the third parties, i.e. the asylum seekers who are forced to move to different locations.

This study shows that the parties become highly dependent on each other to manage the crisis and that this changes the demands made to become more dynamic. Crisis management makes it more difficult to monitor the agent’s hidden actions. Likewise, the principal has difficulty gathering information, gathering knowledge, and exercising control. Procurement of large volumes in a short period of time makes it difficult to build trust. It is also difficult to terminate contracts when there are no alternatives. The study shows that several of the theoretical starting points need to be adjusted to also be valid in this type of situation. The results show even that a crisis can have an impact on sourcing management and on the relationship between principal and agents. Even if the phenomena in phases 1–4 may be found during normal circumstances the crisis cause an exaggerated and dynamic situation of unpredictability. The crisis influences the power balance between actors, and actors’ inclination to act opportunistically (Foss & Stea, 2014). In the pre-crisis situation, the relationship between the procurer (principal) and the suppliers (agents) is as predicted by agency theory (Williamson1979): often characterized by interdependency, long-term relationships, a low level of conflict and a well-functioning market.

When the situation changes from a pre-crisis to a crisis situation, the balance of power may shift in favor of the suppliers as the need for their services and resources dramatically increases, with a simultaneous decrease in the procurer’s ability and motivation to control its suppliers, with the consequence that suppliers act opportunistically, taking advantage of their favorable market position by increasing prices (Hold-up problem) and compromising on their obligation to fulfill the demands in the contracts (Moral hazard). This raises the level of conflict between procurer and suppliers even further. Once the crisis de-escalates, the procurer’s need for its suppliers’ services decreases, and the power balance may tilt back in favor of the procurer. This in turn incentivizes the procurer to act opportunistically and cut costs by terminating contracts prematurely and pressing more forcefully for contracts to be fulfilled (Ellram et al., 2008; Lacity et al., 2008).

Agency theory suggests that information (Hidden information) is shared asymmetrically in favor of the agent, and that the benefits of trust are challenged by the unilateral risk of agents acting opportunistically (Eisenhardt, 1989). However, the scenario sketched above indicates that both principals and agents in a sourcing process can take advantage of their power position and act opportunistically, although the actual form and types of opportunism exercised by the actors are different. While the agents (the suppliers) acted opportunistically by withholding information, increasing prices and not always fulfilling the agreed terms in the contracts (which the principal (SMA) more or less accepted during the acute procurement phase), in the latter part of the crisis, the principal exploited its decreasing requirement for accommodation by terminating contracts in advance, and frequently changing demands and the pre-prescribed order in which the contract was set to end.

This finding confirms previous research by, for example, Brown and Potoski (2003) and Damanpour et al. (2020), which studied and emphasized sourcing as a process and as an organizational capability rather than as separate, dichotomous decisions.

However, the process of sourcing and the changing relationship between procurer and suppliers and its implications for opportunistic behavior has not received adequate attention in previous research (Hussey & Jenster, 2003; Michell & Fitzgerald, 1997). According to TCE theory, the reasons that a procurer acts opportunistically are related to three main features of the relationship between the procurer and the suppliers: asset specificity, uncertainty and frequency of exchange (Williamson, 1979). Empirically, previous research has only been able to establish the positive association between one of these three factors, frequency of exchange (i.e. how often a transaction between a procurer and a supplier occurs), and the likelihood of opportunistic behavior on the part of the procurer (Bhattacharya et al., 2015, p. 266). The results of this study show that, rather than being fueled by frequency of exchange, what really drives opportunism during a crisis is the uncertainty of the situation and the changing circumstances. Actors’ opportunism during crises may thus have explanations other than those that pertain under normal conditions (Ellram et al., 2008; Lacity et al., 2008). In times of crisis, the relationship between procurer and suppliers is unstable. The changing needs and demands fostering time-consuming conflicts where short-termism and the actors’ contrasting long-term goals incite opportunism among both procurer and suppliers.

The insight that the context in which sourcing occurs has a profound influence on how opportunistic behavior develops among both procurer and suppliers constitutes an important contribution to agency theory. In the case of the 2015 migration crisis the friction between agent and principal following the crisis can be tracked to a lack of predictability, lack of time and poor communication. The procurers regarded outsourcing as a quick-fix solutions to the housing shortage quality issues while failing to adequately discuss the contractors’ expectations. The preparatory phase was insufficient and the contracting process inadequate. Based on this experience proactive routines for contracting have to be a part of crisis preparedness plans to involve negotiations and urgently paving the way for establishing a common ground avoiding dire surprises.

7. Conclusion

The study’s findings make three important contributions. Firstly, it shows the potential risks and drawbacks of relying on sourcing in managing a crisis. The legal framework for procurement is largely based on experiences of managing public services in normal situations and not tailored for extraordinary events and the ensuing demands of major crises. The implication is that the legal framework regulating public procurements needs to be adjusted to meet the demand that results from a crisis. Secondly, the findings advance our understanding of how relations between procurer and supplier develop during a crisis scenario. In a crisis context, this relationship is dynamic and unstable, increasing the risk of conflicts between actors. The four phases of a crises highlighted in this study provide a more detailed explanation than existing theoretical frameworks about the challenges to expect in each phase of a crisis when outsourcing is a key component of crisis management. Thirdly, the study contributes to agency theory by showing how the instability of a crisis situation and the relatively short-term perspectives of the relationship between principals and agents influence actors to act opportunistically.

These findings highlight the need for further empirical research into the reasons and extent of opportunistic behavior among public procurers, an aspect largely overlooked in prior studies. Additionally, comparative studies exploring how various EU countries, sharing a common legal framework for public procurements, managed the 2015–2016 migration crisis would enrich the research field. Another highly relevant area for future research would be to conduct more case studies that construct a comprehensive overview of challenges and outcomes associated with sourcing management in times of crisis.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The workd was supported by the Swedish Research Council [grant no: 2016-01680].

Notes on contributors

Magnus Jansson

Magnus Jansson is a trained psychologist, researcher and lecturer in social psychology. Jansson research have mainly been in the field of organizational psychology, behavioral economics and consumer psychology. He currently holds a position as lecturer in social psychology at the University of Skövde and as a senior researcher at the Gothenburg Research Institute at the School of Business, Economics and Law in Gothenburg (Sweden).

David Karlsson

David Karlsson is a Professor in public administration at the University of Gothenburg, Sweden. His research primarily focuses on local government studies, specifically politics and administration of municipalities with a special focus on the actors and institutions of local democracy. He has been responsible for multiple surveys targeting politicians (at all levels of society), citizens, and civil servants. Themes addressed in his research include: representation and representativeness, self-governance, equality and equity, power, political conflicts, outsourcing and backsourcing, issues of administrative division, location politics, and culture policy.

Eric Carlström

Eric Carlström is full professor in Health Care Sciences specialized in management. He earned his Phd in Public Administration. Eric have a former career as manager in the health care industry and have accordingly more than 30 years of experience from ambulance services. He is currently researcher at the Prehospital Disaster and Medicine Center in Western Sweden and University of Gothenburg Centre for Person-centered Care.

Johan Berlin

Johan Berlin is a Professor in Business Administration - public organization. Berlins research is about how and why public organizations bring back previously outsourced work to their own organization (backsourcing). He has also research about collaboration between public organizations, teamwork between professions, management of large organizations and leadership in challenging contexts.

Appendix A.

Interview guide

1. Tell me briefly about your background and role in the organization?

2. Tell me about your experience of outsourcing and backsourcing?

   The refugee crisis 2015–2016

3. What were the reasons why the accommodation services were initially procured from private suppliers?

4. What were the arguments for and against both outsourcing and subsequently backsourcing of accommodation services?

5. What documentation did the organization produce prior to the decision? Could you share it with me?

6. Tell us about how the backsourcing was carried out?

7. Tell us about how you collaborate with suppliers during the backsourcing process?

8. Did you learn anything from the backsourcing process?

9. How did you evaluate the sourcing process?

10. Did it turn out as you expected? What deviated from your expectations?

11. What were the benefits of re-integrating the services?

12. What were the disadvantages of re-integrating the services?

13. Has the termination of contracts and re-integration had any effect on third parties?

14. In retrospect, how do you view the decision to respectively outsource and backsource accommodation?

15. Will you procure accommodation again? What is the long-term strategy?

16. Are there any other aspects or issues that you would like to highlight and tell me about that were not addressed during the interview?

Appendix B.

Subjects interviewed

Table

Interview subject (title)	Organisation	Gender M/W	Pages transcribed	Interview length (min)	ID no.
High official	Justice department	M	12	28	1
High official	Justice department	W	14	34	2
Desk officer	Justice department	W	15	34	3
Section manager	Swedish Migration Agency	M	21	41	4
Section manager	Swedish Migration Agency	M	14	32	5
Section manager	Swedish Migration Agency	M	21	40	6
Section manager	Swedish Migration Agency	M	23	56	7
Unit manager	Swedish Migration Agency	W	18	38	8
Unit manager	Swedish Migration Agency	M	16	32	9
Unit manager	Swedish Migration Agency	M	13	29	10
Unit manager	Swedish Migration Agency	W	15	27	11
Unit manager	Swedish Migration Agency	W	13	29	12
Manager	Accommodation company	M	17	37	13
Manager	Accommodation company	W	17	38	14
Manager	Accommodation company	M	16	35	15
Manager	Accommodation company	M	13	27	16
Manager	Accommodation company	W	10	26	17
Manager	Accommodation company	M	11	25	18
Asylum seeker 1		M	16	30	19
Asylum seeker 2		W	12	27	20
Asylum seeker 3		W	10	25	21
Asylum seeker 4		M	13	28	22
Asylum seeker 5		M	11	26	23
Asylum seeker 6		W	10	27	24
	Total	14/10	353 pages	771 min	24