A voice for the silent: uncovering service exclusion practices

Abstract

Purpose

This paper aims to provide an in-depth conceptualization of service exclusion by drawing on our exploratory research as well as thick and rich insights from the authors’ qualitative data.

Design/methodology/approach

Qualitative research was used to explore service exclusion practices against customers experiencing vulnerabilities. A total of 28 semi-structured in-depth interviews were conducted with refugees residing within Malaysia. The Gioia methodology was used for the authors’ data analysis and the findings were validated by an independent moderator.

Findings

The authors’ empirical findings challenge how service exclusion is currently understood, by adding substantial depth and complexity beyond simply describing “the lack of access to services”. The authors also offer rich empirical findings describing 29 forms of exclusion, which were further reduced to seven types of service exclusion practices: discrimination, restriction, cost barriers, language and technology barriers, poor servicing, non-accountability and non-inclusivity.

Originality/value

This study conceptualizes service exclusion from a process perspective, that is, “how” customers experiencing vulnerabilities are being excluded, rather than “what” is excluded.

Citation

Ng, S.C., Chuah, H.Y. and Nungsari, M. (2022), "A voice for the silent: uncovering service exclusion practices", Journal of Services Marketing, Vol. 36 No. 7, pp. 991-1005. https://doi.org/10.1108/JSM-10-2021-0368

Introduction

This paper is a detailed exploration of the realities of customers experiencing vulnerabilities when encountering various types of service exclusion. In particular, we focus on the plight of refugees, a large population with limited access to social, cultural and economic resources, which can intensify their vulnerability (Fisk et al., 2018; Lee et al., 1999; Shultz and Holbrook, 2009). It has been reported that there are more than 82 million refugees and displaced people worldwide, the highest number since the Second World War (International Rescue Committee, 2022). Recently, there have been increasing calls to alleviate their suffering through the provision of services and designing for inclusion (Boenigk et al., 2021b; Cheung and McColl-Kennedy, 2019; Finsterwalder et al., 2021; Fisk et al., 2018). This is because refugees may be more susceptible to discrimination from service providers and negligence of services (Anderson et al., 2013; Rosenbaum et al., 2017). While the issue of service exclusion has been raised, limited progress has been made to date (Aras et al., 2021).

It has been highlighted that much of the extant literature examining customers with vulnerabilities focuses on identifying issues, rather than providing actionable solutions; that is, very few studies offer solutions that foster inclusivity (Johns and Davey, 2021). Because of this limitation, practitioners may not be able to leverage on sound research insights to take effective action. Thus, we believe that this is an urgent issue that renders immediate attention, especially so since the COVID-19 pandemic has greatly exacerbated the well-being of refugees. Indeed, the ability to achieve well-being outcomes is often dependent on how well services are designed, yet there remains very scarce information on how services should be designed to help customers with vulnerabilities derive value (Nasr and Fisk, 2019). Hence, our study responds to the call by Nasr and Fisk (2019) to further research in this area.

To do so, we first need to be able to identify and understand service exclusion to design such practices out of the service. Yet, to our knowledge, service exclusion is rarely examined in detail as a core construct, but is mostly nested in the literature pertaining to service inclusion (Fisk et al., 2018; Finsterwalder et al., 2021; Leino et al., 2021). Present discussions commonly relate to the types of excluded populations (Fisk et al., 2018), the types of services that customers are excluded from (Alrawadieh et al., 2019; Gordon et al., 2000; Levitas et al., 2007; Saunders, 2008). However, because services tend to be elaborate in nature, the nuances of service exclusion throughout the service delivery process and across the entire service ecosystem should be explored in greater depth to reveal their complexities and impacts. More importantly, doing so emphasizes the difficulties that customers experiencing vulnerabilities face (Beudaert et al., 2017; Ekström and Hjort, 2009; Hamilton, 2007) and offers more possibilities for service organizations and policymakers to design services that are more inclusive.

As such, this paper challenges the current conception of service exclusion, adding substantial depth and complexity beyond a simple lack of access to services. Specifically, our study conceptualizes service exclusion from a process perspective, that is, “how” customers experiencing vulnerabilities are being excluded, rather than “what” is excluded (the traditional focus of studies). Thus, this research makes theoretical and practical contributions by:

• providing an in-depth conceptualization of service exclusion;

• providing rich empirical findings describing 29 forms of exclusion encountered by customers experiencing vulnerabilities, which were further reduced to seven types of service exclusion practices; and

• detailing how such practices are enforced by various actors across the service ecosystem.

In doing so, we respond to the call by Nasr and Fisk (2019) to understand how services can be designed to help these customers derive value. We first review the literature on service exclusion and customer vulnerability because of one’s refugee status and then discuss the methods used in our study. Empirical findings to our research objectives are then presented, followed by a discussion of managerial and theoretical implications.

Literature review

Service exclusion

Service exclusion is broadly understood as a lack of access to services (Gordon et al., 2000; Levitas et al., 2007; Saunders, 2008). From a marketing perspective, it has been defined as the “unfairness that occurs when services (service providers or service systems) deliberately or unintentionally fail to include, or to adequately serve customers in a fair manner” (Fisk et al., 2018, p. 838). To date, service exclusion has been investigated with respect to five main categories (Fisk et al., 2018):

1. marketplace discrimination based on group-level characteristics (e.g. gender-, age- or disability-related);

2. disadvantaged customers at an individual level (e.g. customers with disabilities, women, ethnic minorities);

3. vulnerable consumers lacking power and control (e.g. the elderly, poor, refugees);

4. captive consumers lacking service options other than the current provider; and

5. customers with multiple disadvantages (i.e. intersectional structuring, which includes multiple contexts of gender, educational, religious and racial inequalities preventing or limiting access to services).

Such discussions relate mostly to: the types of excluded populations and the types of services that customers are excluded from (Gordon et al., 2000; Levitas et al., 2007; Saunders, 2008) – which is often presented as obstacles to refugee integration (Alrawadieh et al., 2019; Shneikat and Alrawadieh, 2019; McIntosh and Cockburn-Wooten, 2019).

Further to this, studies have also sought to understand lived experiences of customers facing the lack of access to services. For instance, retail rejection is commonly studied as a form of exclusion in the retail setting, which non-vulnerable customers experience (Ward and Dahl, 2014). Yet, research on service exclusion is often linked to customers experiencing vulnerabilities, because they are more prone to discrimination from service providers and negligence of services (Anderson et al., 2013; Rosenbaum et al., 2017). In the context of a nursing home, the unfair treatment of a patient (primary customer) who is denied access to the service may lead to the unfair treatment and exclusion of the patient’s family members (secondary customer), as they are left without support from the service provider (Leino et al., 2021). Thus, service exclusion not only affects the focal actor but also has significant implications on other actors who are closely related to them.

Nonetheless, service exclusion can go beyond simply “the lack of access to services”, given that services tend to be elaborate in nature. That is, while service exclusion has typically been posited as occurring at the onset of customer journeys (Boenigk et al., 2021b), we argue that it may also happen episodically during the service delivery process of a service exchange that a customer is eligible for. Given that resources can be integrated, used and shared by actors (or individuals) during a service exchange (Finsterwalder et al., 2021; Chen et al., 2020), service exclusion can occur when an actor and/or process prevents access or limits the use of the service by customers, therefore exacerbating their resource condition by inhibiting the transfer of resources. In the context of refugees, they may become more vulnerable given that individuals lacking resources tend to be more susceptible to resource loss while being less capable of gaining resources (Hobfoll et al., 2018). Moreover, resource loss tends to be more powerful than resource gain in magnitude and can affect individuals more rapidly (Hobfoll et al., 2018), thus placing refugees in a worse predicament.

It is also important to consider service exclusion holistically from a service ecosystem perspective, because refugees are often placed into long-term, elaborate service systems which can nevertheless be hostile and restrictive, often obstructing the free flow of actors and resources that co-create value and their well-being (Boenigk et al., 2021a). It also involves numerous actors that mutually influence one another, causing ripple effects across the entire service ecosystem (Farquhar and Robson, 2017) which can be conceptualized at micro, meso and macro levels (Frow et al., 2014; Storbacka et al., 2016; Nasr and Fisk, 2019).

The micro level focuses on the interactions between individuals (Frow et al., 2014), including interactions between refugees and individuals from local communities offering services, as well as service employees encountered during the service delivery process. The meso level considers interactions with focal firms, that is, entities guided by sets of rules (Frow et al., 2014), comprising interactions between refugees and local companies as well as schools, hospitals and financial institutions, to name a few. The macro level pertains to interactions at a broader market level (Frow et al., 2014), comprising interactions between refugees and institutions such as the United Nations High Commissioner for Refugees (UNHCR), regional and national governments as well as other international bodies such as the Association of Southeast Asian Nations (Boenigk et al., 2021a). With such an intricate web of interdependent relationships within a service system (Frow et al., 2014; Storbacka et al., 2016), it is inevitable that refugees may be exposed to more episodes of service exclusion practices over a substantial period which magnify their vulnerabilities.

Hence, it is imperative to further explore the nuances of service exclusion occurring throughout the service delivery process and across the service ecosystem to reveal their complexities and impact. That is, we focus on the need to uncover “how” customers experiencing vulnerabilities are being excluded, rather than “what” is excluded, which is commonly studied. Therefore, we argue the need to deepen our current understanding of service exclusion – indeed, by highlighting “exclusion” instead of “inclusion”, we can further emphasize the difficulties that customers experiencing vulnerabilities face (Beudaert et al., 2017; Ekström and Hjort, 2009; Hamilton, 2007) and thus encourage service organizations and policymakers to be more intentional in designing services, whether in terms of removing or minimizing exclusionary practices.

Customer vulnerabilities because of refugee status

There are several conditions that may intensify customer vulnerability during service exchanges because of the status of being a refugee. This includes the dependency on others – for example, the state, families or community arrangements – for their protection or assistance (Black, 1994). Hence, vulnerability is a state emerging from contexts and circumstances (Backer and Mason, 2012). Once refugees depart from their home country, they tend to be constantly at the mercy of and highly dependent on foreign host governments, local communities, non-governmental organizations (NGOs) as well as international bodies. Such conditions may lead to the loss of personal control (Baker et al., 2005) during service exchanges, which may aggravate consumer vulnerability.

Conversely, the “refugee” label may also facilitate a heightened perception of “institutional dependency by locals, such that vulnerabilities may increase through the wider society’s attitudes and practices towards refugees (Black, 1994). That is, when refugees enter into service exchanges while facing some disadvantage, they may be placed in a vulnerable position during the service process and thus experience discriminatory or even predatory actions from service providers (Baker et al., 2005). These may include acts of sabotage from service employees such as slower service, unnecessary demands for paperwork or outright refusal to serve such customers (Kabadayi, 2019; Razum and Bozorgmehr, 2017). Such sabotage can impede customers experiencing vulnerabilities from accessing (Finsterwalder et al., 2021) or using services during the delivery process. In this case, vulnerability can refer to “a state of powerlessness that arises from an imbalance in marketplace interactions” (Baker et al., 2005, p. 134).

It has also been posited that a lack of access to social, cultural and economic resources can increase the vulnerabilities of refugees (Fisk et al., 2018; Lee et al., 1999; Shultz and Holbrook, 2009). Because refugees lack the necessary knowledge, cultural and economic resources to mitigate harm (Cheung and McColl-Kennedy, 2019; Shultz and Holbrook, 2009), they may be “doubly vulnerable” or more vulnerable to exploitation, including in service exchanges. Nonetheless, it is argued that such vulnerabilities can still evolve if there are changes in resources, abilities or circumstances (Shultz and Holbrook, 2009), thus presenting an opportunity for service researchers to improve refugee well-being through service design (Nasr and Fisk, 2019) by removing obstacles to facilitate the transfer of appropriate resources.

Method

The qualitative methodology of our exploratory study enables us to extract “dynamic, experiential processes and the interactive nature of services phenomena” (Gilmore and Carson, 1996, p. 21). The application of an interpretive approach centres on highlighting respondents’ firsthand experiences of being excluded from services and how exclusion intersects with existing vulnerabilities that they face (Merriam, 1998). Given that our intention is to build rather than test theory (Lincoln and Guba, 1985), qualitative data will provide thick (Gilmore and Carson, 1996) and rich (Shah and Corley, 2006) descriptions, thus allowing us to provide a rigorous account of the phenomena being examined.

Research context and country selection.

We leveraged on refugee experiences of being excluded from services in Malaysia during the transition phase of their refugee journey, that is, taking temporary residence in a host country while waiting to be re-settled (Boenigk et al., 2021a). Malaysia is a critical transit point in this journey in Southeast Asia, being one of the region’s largest hosts – as of 2021, Malaysia hosted about 179,830 refugees and asylum seekers registered with the UNHCR (UNHCR Malaysia, 2021). The Malaysian Government has indicated that these refugees will eventually be repatriated or re-settled, serving only as a transit country on humanitarian grounds (Alhadjri, 2020). Given that there is no intention to integrate or assimilate refugees at this point, service exclusion may be heightened and thus refugee well-being may be compromised. As such, there is an opportunity to explore service exclusion within this context.

Notably, given that refugees are regarded as “undocumented migrants” under Immigration Act 1959/63, they are deprived of any legal rights. They also suffer restricted or unaffordable access to public resources and services (Lego, 2018; Nungsari et al., 2020) such as health care, education, housing, banking and public transportation. It is important to note that refugees are granted a degree of mobility and that there are no encampments (Nah, 2010), so most of them can be found in urban areas, living in close contact with the local community and resorting to the informal job market for their livelihoods (Buscher and Heller, 2010; Ghazali et al., 2020; Salim, 2019; Sullivan, 2016). Given the absence of formal legal frameworks and governance mechanisms, the responsibility of providing assistance and protection to refugees tends to fall largely under the local UNHCR office or with civil society organizations (CSOs).

Sampling

Purposive sampling was conducted to understand the refugee experience of being excluded from services. Our participants were recruited through community partners, including refugee community-based organizations and CSOs providing services to the refugee community. Approaching the community through community partners, who also act as gatekeepers, is an effective way of reducing the mistrust of academic research among refugee communities (Watters and Biernacki, 1989). To obtain a sufficient depth of personal experiences, we reached out to refugees from very diverse backgrounds, with different nationalities, ethnic groups and roles in their community (see Table 1).

Further, to acquire a broad range of experiences, we also contacted refugee community leaders representing their community. It is important to understand that this population comprises of subset communities, which are typically segregated by their country of origin or ethnic group. For instance, amongst refugees from Myanmar, the Rohingya and Chin people form separate communities, many of which are further broken down into smaller groups, this time by local residential areas (the Rohingya of Ampang, Kuala Lumpur, are a separate group from the Rohingya in Ipoh, Perak). Within each, there is at least one community leader or representative, who acts as an interlocutor between community members and external parties such as the UNHCR, NGOs and authorities. As such, given their extensive networks and connections with other refugees, our interviews with community leaders provided insights into the community as a whole. The data from these two sources have enabled us to triangulate the findings, thus reducing any possible bias from just using a single source of data.

Data collection

We conducted 28 semi-structured in-depth interviews with refugees residing in Malaysia. Each interview lasted around 50 min; 15 were conducted through voice calls over telephone or using WhatsApp (the main channel of communication amongst refugees), in compliance with national COVID-19 lockdown measures, whereas the rest were conducted physically when these measures were lifted at their preferred locations. The literature suggests that when meeting such hard-to-reach, hidden and vulnerable participants, it is particularly crucial for researchers to meet them at safe spaces of their choice (Watters and Biernacki, 1989; Ellard-Gray et al., 2015). Accordingly, participants requested to meet at cafes, community centres or eateries near their neighbourhoods. Most preferred to be interviewed away from their homes because of privacy issues – public spaces had the advantage of neutrality as well as being more comfortable for them (King et al., 2019).

All interviews were conducted in the English and Malay languages by one of the authors, except for four Rohingya-speaking interviewees – in these cases, an interpreter was engaged. This was done to reduce the mistrust that refugees might have of interpreters, especially if they are both from the same community (Le Goff and Carbonel, 2020; Refugee Advice Centre, 2010, p. 10). Because of this particular challenge, English- or Malay-speaking participants were mostly recruited. It is important to note that collecting primary data with customers experiencing vulnerability requires a significant amount of rapport-building to ensure trust, because participants may not feel comfortable revealing information that they deem to be sensitive.

Interview protocol

A semi-structured interview guide was developed to facilitate consistency, provide a wider scope for data collection (Corley and Gioia, 2004) and provide a direction for the interview (King et al., 2019). To address our research objectives, we prepared background and experiential questions pertaining to the refugees’ current lifestyles in relation to four broad areas:

1. health and wellness;

2. social situation;

3. financial situation; and

4. security (King et al., 2019).

For instance, we asked refugees about the state of their health and families, as well as how they sought medical treatment. We also asked about their financial situation, whether they were still employed, how they were coping and whether their communities helped. In terms of social situation and security, we had questions about local sentiments and how they were treated by local governments and organizations. The community leaders (also refugees but with leadership roles) were asked similar questions, but additional probes were included in relation to their respective community. Because the interviews took place during the COVID-19 pandemic, related probes pertaining to the pandemic were added to the interviews. This semi-structured format allowed sufficient opportunities for interviewees to bring up unanticipated yet key research perspectives (King et al., 2019), allowing them to discuss their needs and the services that they used to meet them organically.

Data analysis

The interviews were transcribed and shared amongst the first two authors, who read them repeatedly and in parallel to understand the different realities experienced. Using the Gioia methodology, we processed the data individually using open coding to identify first-order concepts which captured the original terminology as intended and maintain a level of scientific theorization (Gioia et al., 2013). We then engaged multiple realities in the natural paradigm to make sense of them and derive patterns that could be repeated (Glaser and Strauss, 1974). Using the Lincoln and Guba (1985) constant comparative method, a conventional match-and-contrast approach, we identified 29 forms of service exclusion, which were further reduced to seven third-order themes – that is, seven types of service exclusion practices. This process enabled us to reduce large textual data to core categories effectively. The Gioia methodology also enabled us to ensure the credibility of our findings (Gummesson, 2005; Seger-Guttmann et al., 2021). Specifically, subjectivity and reflectivity were achieved because data interpretation was derived directly from the participants’ interviews, as expressed by the participants (Crick, 2021).

Further, given that service exclusion exists within the broader context of the service ecosystem, we extracted original quotes from the transcripts to map the occurrence of exclusionary practices across the micro, meso and macro levels to uncover their complexities and nuances. A full list of illustrative quotes is provided in Table 2 to substantiate these insights. In the last stages of data analysis, an independent moderator from one of the local refugee communities was brought in during the peer debriefing process to ensure the trustworthiness of the findings (Corley and Gioia, 2004). In general, these findings were validated by the moderator’s lived experiences.

Findings

In this section, we conceptualize the experiences that customers with vulnerabilities face during service exclusion. Drawing on our interviews with the refugees and community leaders, seven types of service exclusion practices emerged based on 29 forms of service exclusion, as summarized in Figure 1:

1. discrimination;

2. restriction;

3. cost barriers;

4. language and technology barriers;

5. poor servicing;

6. non-accountability; and

7. non-inclusivity.

We also provide an in-depth illustration of how various actors can enforce service exclusion across the service ecosystem. Each type of practice is discussed in detail next.

Discrimination

The unjust treatment initiated by service providers that customers face can occur across three levels of the service ecosystem. At the micro level, refugees may be excluded from local services provided by individuals, such as housing rental or employment opportunities. Such sentiments may stem from personal biases commonly fuelled by xenophobia (a fear or hatred of foreigners in general) and racism (a prejudice against a person’s race, colour or language). Moreover, the misunderstanding of government policies may also encourage individuals to take the safest yet extreme most route by excluding refugees altogether. According to (17):

Renting the house is very […] problematic. Actually [the] government didn't mention the refugees. Government's target is [the] undocumented. Refugees have documents. But some owners they don't understand and they kick out the refugees. This is the main problem.

At the meso level, refugees experience a variety of discriminatory practices by local companies, which may prevent their access to services to which they are entitled. For instance, refugees are sometimes denied legitimate discounts in hospitals or charities may withdraw their support due to negative sentiments expressed by the local community. Therefore, while the government may put policies in place to assist refugees, their implementation may be hindered by public or private companies at the meso level, and thus the potential value of such policies may not reach refugees effectively. Furthermore, while it is under the purview of companies to decide if they would like to offer their services or assistance to refugees, such decisions often hinge on local sentiment as well as those of the government. The above highlights the interconnectedness between actors across different levels of the ecosystem.

At the macro level, discrimination may occur because of government policies or statements, such as not offering legal protection, segregating schools between locals and refugees or even targeting certain ethnic groups (such as the Rohingyas). These actions may be fuelled by the perception of transit or host countries that the arrival of refugees would inevitably strain the resources available to locals (Finsterwalder et al., 2021). Nonetheless, given the influence that the government has on companies and individuals, its attitude plays a large role in shaping the behaviours of downstream actors at the micro and meso levels. This highlights the importance of the government’s role in advocating for the proper treatment of refugees to reduce service exclusion.

Restrictions

Restrictions are limits or controls enforced by service providers experienced by customers. This tends to occur at the meso and macro levels. At the meso level, refugees have access to a limited range of service offerings, for instance, mental health services are rarely accessible, or they face quotas on services legally available to them. At the macro level, refugees may have to deal with fast-changing government policies that not only affect their livelihoods but also affect the manner in which they are portrayed within the host country, in turn directly determining how they will be treated.

Moreover, in spite of having to spend years on end in their host country, refugees are unlikely to receive state support to set up their own establishments to meet their basic needs, even if they have the means or resources. As explained by (18):

I say we don’t have any fund, any help from people. Nobody help[s] here. We don’t know anyone. Just we open by community support […] I also tried to register with UNHCR or government but refugee[s] cannot register any school “cause we are refugee.”

As was the case with discriminatory practices, such behaviours may be because of the need to or mindset of protecting or reserving resources and opportunities for locals. But unlike discrimination, restrictions are often formally imposed by service providers, that is, as a form of systemic bias.

Cost barriers

These include the high costs set by service providers relative to customer income levels. This tends to occur at the meso level, where refugees are often not able to access desired services because of these high costs, whether deliberately or unintentionally set by the service provider. For instance, although refugees can receive a 50 discount off foreign citizens’ rates in accessing public health-care services, many refugees are still unable to afford such essential services because of their lack of sustainable income, as consistent with Chuah et al. (2018). Yet, there remains a lack of cheaper and effective medical alternatives. As (16) notes:

So still he has been using traditional medication. He is scared too […of] not going to hospital. Because he doesn't have money, he doesn't have idea how to cope with this problem.

Similarly for education, there is a strong desire by refugees to send their children to schools, yet many are still not able to afford the school fees. These children suffer, as a result, with some experiencing depression or engaging in illegal activities. Whether deliberately or not, such cost barriers effectively prevent refugees from accessing services that they are eligible for.

Language and technology barriers

These pertain to the preferred language(s) and/or used in the service delivery process that is incompatible with customers, therefore compromising their ability to understand, communicate and use the service effectively. Refugees are likely subject to language barriers at a micro level, as well as language and technology barriers on the meso and macro levels. At the micro level, their inability to communicate with service employees is a common challenge for refugees, who may not speak the main language(s) of their host countries and thus choose to opt out of using the service.

At the meso level, refugees can struggle with accessing services, because companies often do not offer language and technological assistance for the illiterate. As (16) explains:

Most of my people are uneducated. They don't read and write. If they want to go to hospital, they need to scan [QR codes for contact tracing measures]. They need to write their name, they need to write down their telephone number, but they don't know [about these requirements]. So, there is no any helping guide to help them. Cannot go.

At the macro level, refugees may not be able to receive or retrieve information from the government or international bodies easily, again because of language barriers and/or technological barriers (in terms of accessing online communication channels). They are not able to participate effectively or complete the service delivery process successfully, even if they are eligible for such services.

Poor servicing

This item refers to poor service that customers receive as part of the service delivery process. Refugees encounter this sort of service exclusion across all three levels of service ecosystem. At the micro level, refugees may face hostile interactions with service employees, including verbal and physical assault in some cases. According to (12):

Their door [guards] will kick you and their guards are very rude. They will shout even on pregnant women and babies.

At the meso level, refugees can suffer from a lack of professionalism from the company offering the service. As (16) describes:

The clinic normally [is] very, very cold right now. They don't like to talk to the people. They don't like to talk. Symptoms like this, given two to three paracetamol [tablets], one antihistamine [tablet] and then go […] So I can say [it is] “paracetamol hospital”. Nothing else […] Nothing inform to the patients [about any health issues]. Just like this.

At the macro level, refugees may encounter inefficient service delivery processes by local government or international bodies, such as:

• slow and inefficient processing;

• inaccessibility;

• a lack of access to resources; and

• a lack of sound advice.

As such, poor service permeating this ecosystem could make refugees more susceptible to negative experiences, having to cope with unwanted or stressful situations (Morgan and Rao, 2006; Miller et al., 2009). Compounding this is the lack of feedback channels that are usually made available to mainstream customers, which may worsen the service experience for refugees, because there are no outlets for them to make complains or engage in whistleblowing.

Non-accountability

This is with regards to the irresponsible actions the irresponsible actions taken by service providers that customers face during service delivery. Such behaviour tends to occur at the macro level, where local governments or international bodies do not respond to customers adequately. Refugees face a lack of updates and follow-up from their service provider, as well as a lack of transparency in terms of the status of a case(s) in progress. As explained by (15):

Now I'm [waiting] almost four years. No[t] any response from or any results or anything. I don't know where is my profile in UNHCR [its processing system]. Still in RSD [the refugee status determination stage] or still I am asylum seekers or I'm refugees? I am what, I don't know. I don't know my situation, you know?

In most cases, especially with regard to essential services, refugees do not have access to alternative service providers and thus experience service captivity, where their vulnerability is intensified given that the basic expectations of acceptable service delivery such as responsiveness and customer-centric processes are compromised (Rayburn et al., 2020).

Non-inclusivity

Non-inclusivity relates to service providers intentionally making decisions on behalf of their customers, without consulting them directly. Such practices tend to occur at the meso and macro levels. At the meso level, refugees may not have proper representation at the grassroots level. As such, this may compromise the community’s interest by not addressing their needs appropriately. At the macro level, our findings also indicate that refugees are often “not heard”, in terms of having the opportunity to voice their thoughts and opinions. As (17) mentions:

They put their decision over the refugees. Refugees are not given any chance in the decision. This is inhumane. Refugees would [sic, should] have rights to make their own decision.

Decisions are often made on their behalf by local governments or international bodies, without getting discussions or input. Such actions may limit the potential contributions by refugees to the local communities, while reducing their ability to be less dependent on host countries. This inevitably affects the level of perceived control that the refugees have over their lives and by extension, their well-being.

As such, our empirical findings challenge how service exclusion is currently understood, to show how it can go beyond simply “the lack of access to services”. Indeed, customers experiencing vulnerabilities can be and are typically excluded at the onset of services. This can be seen from several practices – that is, discrimination; restrictions; cost barriers; as well as language and technology barriers – imposed by service providers, all of which function to prevent access to the service (see Figure 1). Such practices can deliberately and forcefully impede customers from accessing the services altogether or at least unintentionally result in customers excluding themselves from using such services on their own.

Our findings therefore extend the scope of existing literature by indicating that while customers experiencing vulnerabilities are eligible for services, they can still be excluded at any point during service delivery by actor(s) and/or process(es), whether deliberately or unintentionally, where these customers suffer from limited usage of the service because of:

• poor servicing;

• non-accountability; and

• non-inclusivity.

Furthermore, our findings also reveal how such exclusionary practices are enforced by various actors across the service ecosystem. Hence, we posit a more holistic definition of service exclusion, which is as follows: the process of preventing customers from accessing and/or limiting the use of a service through actors and/or processes within the service ecosystem, whether deliberately or unintentionally. By doing so, we provide some conceptual guidance to scholars studying service provisions, in particular service exclusion.

Discussion

Theoretical and managerial implications

This paper provides an in-depth conceptualization of service exclusion, drawing on exploratory research as well as thick and rich insights from our qualitative data. By doing so, we achieved our intention of building relevant theory underlying this concept. To date, our knowledge and research on service exclusion remains superficial and relate mostly to:

• the types of populations that are excluded (Fisk et al., 2018); or

• the types of services that customers are excluded from (Alrawadieh et al., 2019; Gordon et al., 2000; Levitas et al., 2007; Saunders, 2008).

In contrast, this study conceptualizes service exclusion from a process perspective, that is, “how” customers experiencing vulnerabilities are being excluded, rather than “what” is excluded (the traditional focus of studies). This approach enables us to highlight the extent of the difficulties that each customer with vulnerabilities may face (Beudaert et al., 2017; Ekström and Hjort, 2009; Hamilton, 2007), instead of reporting them merely as statistics.

Our empirical findings challenge the current conception of service exclusion, adding substantial depth and complexity beyond a simple lack of access to services. Specifically, our findings reveal that service exclusion can occur:

• at any point during the service delivery process, rather than just at the onset of a customer journey;

• when an actor and/or process prevents access to or limits the use of the service by refugees, whether deliberately or unintentionally, thus inhibiting the transfer of resources;

• across the service ecosystem, as enforced by various actors; and

• occur concurrently with other service exclusion practices, rather than in silos or a linear fashion.

For instance, in a health-care setting, refugees can experience language barriers and poor servicing. Hence, to expand upon the definition offered by Fisk et al. (2018), we extend current literature by re-defining service exclusion as: the process of preventing customers from accessing and/or limiting the use of a service through actors and/or processes within the service ecosystem, whether deliberately or unintentionally.

This paper also offers rich empirical findings describing 29 forms of exclusion, which were further reduced to seven types of service exclusion practices:

1. discrimination;

2. restriction;

3. cost barriers;

4. language and technology barriers;

5. poor servicing;

6. non-accountability; and

7. non-inclusivity.

We argue that this knowledge is fundamental for effectively designing services that will actually be of value to and user friendly to customers experiencing vulnerabilities. Indeed, the ability to achieve well-being outcomes is often dependent on how well services are designed, yet there remains very scarce information on how services should be designed to help customers derive value (Nasr and Fisk, 2019). As such, our study responds to the call by Nasr and Fisk (2019) to further research in this area. Practitioners can use our framework to identify and pre-empt service exclusion practices by evaluating various potentially non-inclusive aspects of services, thus enabling them to design these mindfully.

In particular, we highlight three practices which can be categorised as limiting the usage of service:

1. poor servicing;

2. non-accountability; and

3. non-inclusivity (see Figure 1).

From a service provider’s perspective, these tend to be less obvious than those preventing access to services and are often overlooked. Nevertheless, such practices are often discussed and highlighted by refugees because of their lived experiences from direct interactions – unless one experiences such conditions personally, it is practically impossible to know of their existence and extent. In this regard, perhaps more education and training could be done to enable further empathic interaction with refugees. Some potential steps for practical interventions include implicit bias training, education on cultural sensitivity and inducing or requiring volunteer work with vulnerable communities as part of workplace corporate social responsibility or diversity, equity and inclusion initiatives. Not only can these positively affect the reputation of companies and improve their service culture but also empower them to positively affect the experiences of customers with vulnerabilities seeking services.

Nevertheless, our results also reveal that eradicating some service exclusion practices may not be entirely straightforward – these tend to require some form of collective effort or coordination from actors across the ecosystem for successful implementation. For instance, while the government may put health-care policies into place to improve the well-being of refugees at the macro level, local hospitals may choose to deny them the legitimate discounts that they are entitled to. On the other hand, government policies can also be enforced or retracted because of local sentiments. Thus, the implementation of government measures can be hindered by public or private companies operating at the meso level or even individuals at the micro level, such that the potential value of these policies may not reach refugees effectively.

Moreover, efforts made by companies and organizations to improve the well-being of customers experiencing vulnerabilities can also be impeded by local sentiments or the government. For instance, a business donor withdrew its support for the refugee community because of backlash from the local community. Misunderstandings of government policies encouraged individuals to take the safest yet most extreme route by excluding refugees from services altogether. The above complications highlight the interconnectedness and interplay of actors across different levels of the ecosystem in affecting the provision of services. Depending on how these actors work together, they may co-create or co-destroy value for customers (Echeverri and Skålén, 2011; Ng et al., 2016, 2019). Therefore, strategies aimed at tackling these practices should be holistic enough to address various actors across the ecosystem, to encourage the successful implementations of strategies.

Future research agenda

There are two areas of future research to consider. The first is the potential effects of service exclusion on customers experiencing vulnerabilities, which may be grossly underestimated. We posit that refugee vulnerabilities can be magnified by constant exposure to episodes of service exclusion practices over a substantial period of time. Short-term exposure may lead to negative emotional states, whereas long-term exposure may lead to more negative outlooks toward life and their loss of trust in institutions (governments and global bodies). The latter can lead to severe mental and associated health issues. Yet, health care related to mental health is often not accessible, because such services are considered non-essential. According to (1):

Those NGO I refer her [to them] but unfortunately no one accepts her. Still she's looking for a place. Even hospital didn't accept because they say we won't take any new case of […] mental issues.

As such, it is critical for future research to shed light in this area, so that host countries and global bodies can alleviate suffering.

Second, current research is on a positive trajectory in terms of promoting service inclusion, from the service providers’ perspective. For instance, the four pillars of service inclusion:

1. promoting enabling opportunities;

2. offering choices;

3. relieving suffering; and

4. fostering happiness (Fisk et al., 2018).

But while refugees may be eligible for these services, many may not be able to participate effectively to complete the service delivery process. As the saying goes, you can lead a horse to water, but you can't make him drink it. Therefore, more research can be done in terms of understanding how refugees view the current services available to them and how we can shape the attitudes of customers with vulnerabilities so that they can be more positively inclined toward participating in such services, as well as co-creating them with service providers. Such activities often require an extensive amount of upskilling and upgrading efforts on the part of refugees, especially if they are illiterate. Moreover, refugees may have a more pessimistic and depressing outlook toward life, which will make this task harder. As such, more knowledge in this area is needed to address such concerns.

Lastly, we have a word of advice for researchers approaching refugees and customers experiencing vulnerabilities that would be meeting them in their own safe spaces, which include considerations of actual physical spaces (the location and mode of meeting) and the manner in which the interview is set up (the types of languages used, the speed of interviews). Researchers should not rush the data collection process with customers with vulnerabilities, because many have typically had past traumatic experiences that could make responding to probing questions and constant interrogations during the data collection process very difficult.

Conclusion

Our intent is to lend our voices to this silent population, by extending an opportunity for them to narrate their own intimate encounters with service exclusion. Through detailed insights gathered from firsthand experiences, we are able to challenge and re-define what service exclusion means for customers experiencing vulnerabilities. We hope that academics and practitioners alike can make use of relevant findings to relieve the suffering of refugees and to facilitate service inclusion.

Figures

Figure 1

Summary of service exclusion practices

Table 1

Profiles of interviewees

Code	Ethnicity/nationality	Time in Malaysia (years)	Residence in Malaysia	Status	Community leader
01	Afghan	7	Kuala Lumpur	UN card	Yes
02	Kachin (Myanmar)	10	Kuala Lumpur	UN card	Yes
03	Karen (Myanmar)	>10	Kuala Lumpur	UN card	Yes
04	Mizo (Myanmar)	>9	Kuala Lumpur	UN card	Yes
05	Mon (Myanmar)	20	Kuala Lumpur	UN card	Yes
06	Rohingya (Myanmar)	>20	Kuala Lumpur	UN card	Yes
07	Rohingya (Myanmar)	7	Perak	UN card	Yes
08	Rohingya (Myanmar)	7	Kuala Lumpur	UN card	Yes
09	Somali	5	Kuala Lumpur	UN card	Yes
10	Syrian	6	Kuala Lumpur	Mahar card	Yes
11	Palestinian	9	Selangor	Work permit	Yes
12	Syrian	6	Kuala Lumpur	UN card	No
13	Syrian	2	Kuala Lumpur	Spouse visa	Yes
14	Syrian	>3	Kuala Lumpur	UN card	No
15	Syrian	>3	Kuala Lumpur	UN card	No
16	Rohingya (Myanmar)	>7	Kuala Lumpur	UN card	No
17	Rohingya (Myanmar)	14	Kuala Lumpur	UN card	Yes
18	Rohingya (Myanmar)	>8	Kedah and Kuala Lumpur	UN card	Yes
19	Rohingya (Myanmar)	8	Kuala Lumpur	UN card	No
20	Rohingya (Myanmar)	5	Kuala Lumpur	UN card	No
21	Rohingya (Myanmar)	10	Kuala Lumpur	UN card	No
22	Rohingya (Myanmar)	5	Kuala Lumpur	UN card	No
23	Sudan	4	Kuala Lumpur	UN card	No
24	Rohingya (Myanmar)	7	Kuala Lumpur	UN card	No
25	Rohingya (Myanmar)	19	Kuala Lumpur	UN card	Yes
26	Rohingya (Myanmar)	26	Terengganu	UN card	Yes
27	Rohingya (Myanmar)	20	Pahang	UN card	Yes
28	Syrian	3	Kuala Lumpur	UN card	No

Table 2

Service exclusion practices across the ecosystem

Service exclusionary practices
	Example quotes
I. Prevent access to service
1. Discrimination	The unjust treatment initiated by service providers that customers face
Micro-level	Hindering leasing and employment opportunities: Renting the house is very.. err.. problematic. Actually government didn't mention the refugees. Government's target is undocumented. Refugees have documents. But some owners they don't understand and they kick out the refugees. This is the main problem. They mentioned the PATI, pendatang tanpa izin [illegal migrants]. Refugees [are] also pendatang tanpa izin [illegal migrants], but they are already allowed to stay on humanitarian ground. So [landlords] could not do this. (17) Very afraid local people. Because local people looking us very badly… They don't want to give Rohingya people working, their plan. They don't want to give open any business. They don't want to give driving any motor or car and especially they don't want Rohingya people in Langkawi. Other place. And some also don't allow to rent the house. (18)
Meso-level	Against specific group within refugees for banking: I'm not sure about Palestinian but if you are Syrian specifically. Yes this is a problem Yemenis have. No. Their nationality no. Iraqi other no. But for Syrian the problem, even if you are legal, I'm not saying illegal. If you're working visa and company and your company have account, Banks A and B [names redacted] close them. The only one allow is Bank C [name redacted]. (12)
	Deny access to legitimate discounts in hospitals: But not yet doing. They say before we have discounts, now we cannot give you discount anything. You have to pay full like private hospitals… Even if we have card, they say like this. (18)
	Stop donations due to pressure from locals: We face some donors who.. like for example one.. there's one company which makes flour for bakery. So they were supporting us at the beginning of MCO. The refugees. Then when that.. when the whole situation and the whole xenophobia started, they said we don't want to put ourselves at risk so we want to stop. Yes. But it doesn't mean all of them… (10)
Macro-level	Segregation between locals and refugees in schools: We are not allowed to study at the local school. At least the government should.. should work together with the NGOs to provide the system, curriculum system of Malaysia. To.. yea.. A lot of things the government can do. Not the.. We are not asking.. not ask to study at the local schools but the schools we have now. (07)
	Against specific group within refugees for employment: Right now we are reported that some employers they ask whether you are Myanmar Muslims or Rohingya or… But especially they targeted the Rohingya. If they say they're Rohingya, so yeah they don't accept the Rohingya. Said that because.. because of the government statement. So the government statement you know it can harm all the Rohingya community here in Malaysia… (08)
	Not offering legal protection: The first one is the legal protection from Malaysian government. For the Rohingya refugees. Even.. we are (not) demanding or ask to the government to ractify 1951 Convention but there should be a law for refugees. Recognise the.. refugees.. not to arrest the refugees… (07)
2. Restrictions	The limits or controls enforced by service providers experienced by customers
Meso-level	Limited range of service offerings: I had a case and that it needs to take medicine. She used to take but unfortunately I.. those NGO I refer her but unfortunately no one accepts her. Still she's looking for a place. Even hospital didn't accept because they say we won't take any new case of.. like mental issues. (01)
	Quota on available services: If you cannot get a space, or a quota for free of charge, then you have to go for private or government hospital. But the government hospital some of our community members said that for government hospital, you know the price is around RM300. But the thing they don't want to test refugees or other foreigner workers. Unless they have symptoms. Because the kits is also like.. Not a lot of machines there you know. (03)
Macro-level	Fast changing government policies: But for those who have UN card also, the owner are scared to.. how to say that.. are scared to get back to the work. Because of the policy changes and a lot of.. how to say that.. sound of the local, media and also like some the government.. voices and all that. (02)
	Restrict access to mosques: You shouldn't put in your statement that foreigners not allowed to go to the mosque. Or Malaysians are priority in this. You shouldn't, you shouldn't, you should just say mosque can… we can go to mosque to perform prayers but due to the COVID, we will limit the space. So maybe instead of 100, we only take 30 people in. You don't need to specify.. you don't need to be very specific about which ethnicity should go in…. (06)
	Lack of support to set up establishments: So suddenly a NGO [name redacted] arrive in Langkawi. How.. how you provide this school and how do you provide for this? I say we don't have any fund, any help from people. Nobody help here. We don't know anyone. Just we open by community support. So they.. can we.. they would like to join. So OK. So we also try to… I also tried to register with UNHCR or government but refugee cannot register any school 'cause we are refugee'. (18)
3. Cost barriers	The high costs set by service providers relative to customer income levels
Meso-level	High costs of hospitalisation/education, relative to refugees' incomes: If we have any kind of disease, we go to clinic they charge us a lot of money. We have to where cost is very low. We have to go there. We can't even go hospital because they charge us very high cost. We can't even go to hospital. (209) So yeah, some some cases of girls. Let's say they are over 17 and they are supposed to go to college or University but they can't afford that. I was reached by a family that's telling me that their daughter is thinking about suicide. Because she's 19, she's been out of study for two years and she thinks that it's it's like worthless. Her life is is meaningless. She doesn't have friends. You can't go out. It was pretty difficult for them… (13)
	Lack of cheaper and effective medical alternatives: … one of the guys got severe motorbike accident… And broken his right leg or left leg. Completely broke. So still he has been using traditional medication. He is scared to.. not going to hospital. Because he doesn't have money, he doesn't have idea how to cope with this problem. I don't think that he will recover with this traditional medication. His leg is broken like this. (16)
	Lack of grants/fundings to finance school fees: Yes I can go. But I have to pay money. I cannot pay. This is for me very big problem. If I said to principal, please try to take some money and discount for me. He said are you refugee? I said yes. He said everyone is refugee, how can I help you. If you cannot pay money, you stop your study, we don't want. He said like that. For me study very important. That's why I'm doing.. still study. (20)
4. Language and technology barriers	The preferred language(s) and/or technology used in the service delivery process that is incompatible with customers, therefore compromising their ability to understand, communicate and use the service effectively
Micro-level (language)	Inability of employees to communicate: But some of the patients they don't know how to speak. Old or young, the mother or father to follow [to hospital]. One of our coordinators has to follow. So 3 persons. Not allowed to get in. (05) They went shopping but without mask. So Company A [name redacted] refused to let them in. They thought it is because they are Rohingya, so they get kicked out. They insist to go in. That person said wear mask, wear mask. They don’t understand and went home. (25)
Meso-level (language and technology)	Lack of language and technology assistance for illiterate: So normally this moment, most of my people are uneducated. They don't read and write. If they want to go to hospital, they need to scan. They need to write their name, they need to write down their telephone number, but they don't know. So there is no any helping guide to help them. Cannot go. (16) Because they think many questions, how can I answer, I don't know both language in Bahasa and also English. Cannot go in. So if they want to go anywhere else, for example, they can come to KL with RM2 - RM2.5. But they hire 1 taxi because they don't know how to scan, how to take bus how to go to train. They hire 1 taxi or grab. Please help me for book Grab and book. So I feel very sad looking. Although they don't have much money, they have to go for any reason. They go Grab. (16)
Macro-level (language and technology)	Ineffective channels of communication: We only read the letter sent by the UN. Just wait for the people. Whether UN is open or not, we just hear from pople. UN has already given their facebook, youtube. But even if they gave us, we don’t know how to read or what to do. Many Rohingya are not educated. (26)
II. Limit usage of service
5. Poor servicing	The poor service that customers receive as part of the service delivery process
Micro-level	Hostile interactions: You know if I go hospital.. so some of the.. like this.. he has wife. His wife is pregnant. Very pain. Ok. Give hospital appointment is so long time. He said my wife will die, so long appointment. And ask me where are you from? I came from Myanmar. They say go your country. (20)
	Verbal and physical assaults: Their door will kick you and their guards are very rude. They will shout even on pregnant women and babies. Yeah. And then there is no online to follow up with your case. (12)
Meso-level	Lack of professionalism: Basically I can say the NGO clinic not proper clinic, I can say paracetamol clinic… Panadol clinic. Nothing else. Nothing else. The clinic normally very very cold right now. They don't like to talk to the people. They don't like to talk. Symptoms like this, given 2-3 paracetamol, 1 antihistamine and then go. Although the patients came to hospital from 7 in the morning and she is going back after 3 o'clock with 2-3 paracetamol. So I can say paracetamol hospital. Nothing else. They only take blood pressure. Go. No good no bad. Nothing inform to the patients. Just like this. (16) The next time when I was sick, one week just we go from hospital to another hospital they sent us because I have UN card. So they said to me that time you go.. It was Wednesday. I went Wednesday. They told me go to the clinic Wednesday. And from my house to the clinic around 40. Very far. I went there. They said to us, we don't.. We are closed Wednesday. How? That day I call the UN who told me go there and give me the location. I asked, it's closed. So what I do? But she cannot do anything. Sorry. That's what she said. (28)
Macro-level	Slow and inefficient processing: Many UNHCR card also.. expiry date already past. Many people also do not have any, already expired. So they cannot extend their card. When UNHCR calls them, they can call. 5 months, 6 months also 7 months, many people waiting. So they.. if police get them, some people sympathize them, some people also don't sympathize, take them directly to the jail. (16) So many problems in terms of documentation because some of the women, pregnant women are about to deliver babies and their husbands actually have documents.. UN documentation. Registered with UN. But they are not. They are still in the process of add on you know. Like say waiting for UNHCR to add on their case with the spouses or something like that. They are still waiting but they don't have the documentation. So it's… it's hard for them also. It's hard for them also to get discount or sometimes. Discount is one thing but the thing is. You cannot go to government hospital. (03)
	Inaccessibility: They don't answer phone, they don't answer email. And they don't allow anyone, the people to come to office. Even when they, after they open. They also keep, don't allow anyone to go there and you want to go you need appointment, call their phone number no answer. (11)
	Lack of access to resources: Some families contacted us. We visited them at home. They had zero food in their fridges. We.. we tried to talk to UNHCR. They told us that they have nothing to provide so we started seeking other. (13) I think it is UNHCR. Every work in the world, they work together with World Food Programme (WFP). And international NGOs. Everyone from every country, everywhere in refugee situation, they are able to enjoy their relieve. Food from international or UN NGOs or international NGOs. But the refugees in Malaysia. (17)
	Lack of sound advice: So this problem I already submit to the UNICEF office and I submit to the government also, special branch that human traffickers finding me to kill. So UNHCR, nothing, reply me. UNHCR said like this, you close your phone and then you change your phone number, you stay in any place, don't go anywhere. What is that? I don't understand they telling me. No help me anything yet. (18)
6. Non-accountability	The irresponsible actions taken by service providers that customers face during service delivery
Macro-level	Lack of transparency: Just. and the problem with the UN also, there is no.. you can't know anything about your case. Like in Turkey, Lebanon, Jordan, we can open online website, your enter.. here's my number or your case now you are here. Here they… Now this year they implemented just to contact us, the simple page, contact us. And the other problem also there is no regulation or rules. So, you know what, like what is my status like you know when you want to immigrate to Canada they have this point if you are married, educated, like point point point. Here nothing. (12)
	Lack of updates and follow up: So this is exactly what, what.. What.. what happened and now I'm almost 4 years. No any response from or any results or anything. I don't know where is my my profile in in in UNHCR. Still in RSD or still I am asylum seekers or I'm refugees? I am what I don't know. I don't know my situation, you know? (15)
7. Non-inclusivity	Service providers intentionally making decisions on behalf of their customers, without consulting them directly
Meso-level	Lack of proper representation at grassroots level: Since I came I found the community leaders same like them. They did not change, so you cannot change them also. Because like appearance they are like like a family like this relatives, they hold the community and they do what they like and they are not helping. They are not helping the community. We do ourselves, sometimes we contact with. We do ourselves. Sometimes we try to help also our community by buying.. individual.. individual like that. Because the community is not taking the role.. the leader role of the.. of the community… I complained, I told them this community, the leaders they are not representing me. Because they have some problems also. They're not taking care about the community as well as they should. (24)
Macro-level	Exclude refugees from decision-making: First class human rights violation goes in hands of UN agencies. They never listen to anyone. They are dictator. They put their decision over the refugees. Refugees are not given any chance in the decision. This is inhumane. Refugees would have rights to make their own decision. (17) We have no other country to go. So we are also like.. whatever the government suggest. or whatever that the government decide for us. We have to be able to accept even though the negative or the positive. (02)