Keywords

Introduction

Global disasters have been increasing in diversity and severity in the past decade (IFRC 2012). To mitigate the effect of such disasters, humanitarian relief organizations (HROs) across the world are busy rescuing and helping people in disaster-prone areas, where the poor infrastructure often makes humanitarian logistics challenging. In such emergency operations, there are multiple stakeholders such as commercial enterprises, donor and host governments, the military, international and local HROs, as well as local communities working together for effective response. These agencies often have varied motivations, mandates, resources, and technical expertise, and must coordinate well to ensure the effectiveness of the emergency operations, as no single actor has sufficient resources to respond effectively to a major disaster. Therefore, coordination in emergency humanitarian logistics is receiving increased attention (Balcik et al. 2010).

Trust is a critical factor for effective coordination in supply network management (Golicic et al. 2003; Gulati and Singh 1998). It is well documented that a high level of trust between supply chain partner organizations leads to coordination effectiveness and better chain performance (Fawcett et al. 2008). However, the literature tends to focus on trust in long-term relationships. Studies on trust in the temporary networks formed by emergency logistics or quick onset disasters are scant (Tatham and Kovács 2010). In the context of humanitarian logistics, these temporary networks are often the norm in emergency relief operations, whereby a number of individual logisticians from a variety of organizations have to work together. Moreover, among the various types of temporary networks, groups in humanitarian logistics operations are better classified as hastily formed networks (HFNs) or emergent response groups (Ben-Shalom et al. 2005; Majchrzak et al. 2007). While sharing a common aim to helping the disaster victims, the logisticians in such groups may have neither worked together before, nor have gone through the same training. An emergent response group may develop, migrate, and reorganize, gaining and losing memberships in an unstructured way (Majchrzak et al. 2007). In such a context, trust building would follow a different pattern from trust in a long-term relationship, which would normally come about through positive past collaboration (Fischer 2013). Individuals within such a network are often tied together via “swift trust” or “initial trust” (Meyerson et al. 1996).

Swift trust is a form of trust occurring in temporary organizational structures, assumed by group members initially, and is later verified and adjusted (Meyerson et al. 1996). It is already recognized as an important type of trust in humanitarian operations (Stephenson 2005). Tatham and Kovács (2010) have proposed a model of swift trust in the humanitarian context with several possible facilitators. However, to date, there is no empirical work on swift trust in the study of humanitarian logistics.

To fill this research gap, we first develop a research framework with testable hypotheses to advance our knowledge in this field. We not only examine the factors affecting the forming of swift trust but also the impact of swift trust on the coordination activities as well as their effectiveness in humanitarian networks with coordination theory as the tool (Malone and Crowston 1994). By applying swift trust and coordination theory into the humanitarian logistics practice, this study would enrich our understanding in this important research field and contribute to the improvement of humanitarian relief operations on the ground. Practical implications from the study may help HROs build trust rapidly and effectively with their unfamiliar partners on-site. Such trust may lead to a smoother and hence more effective coordination and collaboration of their logistics operations, which can in turn relieve the sufferings of the beneficiaries and reduce the HRO operating cost. From the perspective of the research methodology, it is one of the few empirical investigations in the field of humanitarian logistics and may contribute to a more rigorous empirical examination in the field.

Literature Review

Trust is a core concept in supply network management and has been studied from a range of perspectives, including economic, psychological, and sociological. It integrates “micro-level psychological processes and group dynamics with macro-level institutional arrangements” (Rousseau et al. 1998: 393). Morgan and Hunt (1994) generally define trust as “confidence in an exchange partner’s reliability and integrity.” In the humanitarian context with a focus on inter-personal relationships, trust is defined as “a fundamental belief that the other can be relied upon to fulfil their obligations with integrity, and will act in the best interests of the other” (Tatham and Kovács 2010: 37).

According to Stephenson (2005), interorganizational trust in the humanitarian context can be classified into four aspects:

  • Trust based on the judgment of goodwill and how much one considers the other to be a friend (companion)

  • Trust based on the perceived ability of others to carry out the needed tasks or to get the job done (competence)

  • Trust based on whether the behavior is consistent with contractual agreements (commitment)

  • Trust based on expediency because of the need to accomplish the goals quickly (swiftness)

As HFNs are commonly seen in emergency humanitarian logistics operations, swift trust is crucial to HFNs by the achievement of trusting interpersonal relationship in a very short time. Thus, we focus on swift trust in this study.

The concept of swift trust is first proposed by Meyerson et al. (1996). This concept is developed as an explanation of trust development in temporary, nonconventional teams within or between organizations. It is “a unique form of collective perception and relating that is capable of managing issues of vulnerability, uncertainty, risk, and expectations.” Instead of the typical trust built through the passage of time as an evidence-driven information process, swift trust is created by category-driven processes under very tight time constraints. A temporary team would interact as if trust were present, but then must verify that the team can manage the expectations of all stakeholders. It is conditional and is in need of reinforcement and calibration by action. Hung et al. (2004) have developed a framework for the initial formation and further growth of the attribute of swift trust. Three routes to trust, the peripheral, central, and habitual, are proposed in Hung et al.’s framework. The peripheral route refers to the early establishment of trust, and the central one is its further development in relationships with a long-term perspective, while the habitual one is at a higher level where trust is based on patterns developed within the relationship.

In the context of emergency humanitarian logistics operations, networks are often formed with little or no prior warning, and the peripheral route of trust building would be more relevant and important. According to Hung et al. (2004), there are five antecedent conditions influencing the formation of swift trust, namely, third-party information, dispositional trust, rule, category, and role. Tatham and Kovács (2010) further proposed a model for the route to swift trust. Besides the five conditions, the model includes a feedback loop with the communication environment and perceived risk as moderators. In a related discussion, Tatham and Kovács (2010) also explored the importance of third-party information and the development of common rules in humanitarian logistics operations when forming swift trust.

Building trust is not the end in itself. The purpose of forming trust is to facilitate more effective interorganizational coordination and collaboration. The most commonly accepted definition of coordination is “the act of managing dependencies between entities and the joint effort of entities working together towards mutually defined goals” (Malone and Crowston 1994). Based on coordination theory, proposed by Malone and Crowston (1994), we may see that the coordination in humanitarian logistics arises from the task interdependency, where a single entity is unable to meet the needs of the beneficiaries in one location. One coordination mechanism is to assign a single lead agency (often called the umbrella organization) as the coordinator to facilitate horizontal coordination (Akhtar et al. 2012). It is often called the cluster approach and is widely used in humanitarian relief operations (Jahre and Jensen 2010). However, mechanism alone is insufficient for effective coordination as most HROs are independent from one another with different motives; interorganizational trust is essential for effective coordination. In the situation of complex and dynamic environments, a context similar to emergency humanitarian operations, Hossain and Uddin (2012) have developed a framework to model coordination within social networks. Similarly, Saab et al. (2013) have explored the connection between trust and coordination among the field-level Information and Communications Technology (ICT) workers in HROs and shown that trust building through collaborative activities is essential for successful interorganization coordination.

While there is a growing recognition of the importance of trust in humanitarian logistics operations, and specifically swift trust, there is no known empirical study on the forming of swift trust in emergency relief operations as the existing works are largely based on theoretical applications such as Tatham and Kovács (2010) or case studies such as Saab et al. (2013). This study is thus initiated to fill the research gap.

Hypotheses and Framework

To develop a framework with empirically testable hypotheses, we first examine the model presented in Tatham and Kovács (2010: 38). The five antecedent conditions in the model are then further developed into testable hypotheses in the context of emergency relief operations.

Third-party information is the first factor in the original model. It is mainly in the form of reputation or certification from our field study with humanitarian logisticians on the ground. Third-party information can also manifest in the form of reference, particularly for the small local HROs. When a partner is introduced by a trusted person or organization, two HROs without any prior work experience would be easier to trust each other. We thus posit the following hypothesis.

H1: It is easier to generate swift trust toward a partner if the partner is introduced by a trusted person or organization.

The second factor, disposition trust, is not specific to the humanitarian context (Tatham and Kovács 2010), and as such, we do not explore it further. Rule is the third factor, which is important in relief operations. It has been called to develop common rules and procedures among HROs in the same cluster to facilitate interorganizational coordination, though in practice, there is still a long way to go (Balcik et al. 2010). Two HROs with similar operating rules and procedures may generate trust in each other much easier, resulting in the following hypothesis.

H2: Two HROs are more likely to generate swift trust toward each other if they follow similar rules and procedures in their humanitarian operations.

Category is an important though often divisive factor in the forming of swift trust. Here, we specifically explore the impact of organizational belief and value, one important type of category. Sharing the same organizational value or belief would be particularly important for swift trust forming among the small, local HROs. In addition to organizational values, religious belief and conviction could also play an important role in relief operations as many nongovernmental organizations (NGOs) are faith based. Anecdotal information from the field suggests that it is relatively easier for two Christian HROs to work together as compared to a Muslim organization working with a Christian one. We thus have the next two hypotheses.

H3a: Two HROs are more likely to generate swift trust toward each other if they are similar or share the same organizational beliefs.

H3b: Two HROs are more likely to generate swift trust toward each other if they are similar or share the same religious belief or values.

Role is another important factor, which is affected by the competency of a person or an organization. It is reported that experience in prior disaster relief operations is important for trust building (Saab et al. 2013). A person or organization experienced in humanitarian operations may be viewed as being more competent and as such is easier to be trusted, resulting in the following hypothesis.

H4: It is easier to generate swift trust toward a partner if the partner is perceived as being competent by his/her background or experience.

In addition to the five antecedent factors, perceived risk can act as a moderator on swift trust building. In the humanitarian context, the competition for media attention and the subsequent funding would be an important risk as most HROs do not have stable revenues but rely on donations from the goodwill of various individuals and organizations (Tomasini and Van Wassenhove 2009). A sense of the potential competition may affect swift trust building between two HROs and can thus lead to them withholding critical information and deterring coordination effectiveness. Another possible risk is the potential clash due to value differences, which is a common factor in interorganizational conflicts and happens also in humanitarian relief operations (Akhtar et al. 2012). We thus propose the next two hypotheses.

H5a: It is more difficult to generate swift trust toward a partner if the partner is perceived to be a potential competitor for funds.

H5b: It is more difficult to generate swift trust toward a partner if the partner is perceived to be significantly different in organizational values.

In addition to the antecedents of swift trust, we also investigate the impact of swift trust on the coordination activities as well as coordination effectiveness in the humanitarian context. These relationships are well known in the coordination literature (e.g., Gulati and Singh 1998; Rampersad et al. 2010). We now posit the following two hypotheses.

H6: The higher the level of swift trust among the coordinating partners, the higher the degree of their coordination activities.

H7: The higher the degree of coordination activities among the partners, the better the coordination effectiveness.

Figure 9.1 summarizes the six hypotheses into a framework.

Fig. 9.1
figure 1

Research framework

Full size image

Research Sample, Measures, and Results

Research Sample

We choose HROs in Southeast Asia as the research sample. Several countries in the region such as the Philippines, Indonesia, and Myanmar are particularly disaster prone, given the various recent disasters in the region such as tsunamis (2004 Indian Ocean tsunami in Aceh), earthquakes (2009 Sumatra earthquake in Padang), cyclones and typhoons (2008 Cyclone Nargis in Myanmar, 2013 Typhoon Haiyan in Philippines), floods (2011 Bangkok flood), and volcanic eruptions (2010 Merapi eruption in Java). In addition, there are man-made disasters such as ethnic conflicts in Myanmar and Southern Philippines. Many large international HROs have manned active operations in the region to support their extensive relief and development programs. It is, indeed, an appropriate location for our study.

Moreover, most countries in the region are democratic countries with a vibrant private sector and strong networks of local albeit much smaller HROs. For instance, Indonesia, the largest country in the region, is estimated to host tens of thousands of local NGOs, of which at least 9,000 are officially registered with the Ministry of Home Affairs (Figge and Pasandaran 2011). Many NGOs are heavily involved in humanitarian operations and are HROs as well. There are both religious and secular NGOs, those focusing on emergency relief and those having both relief and development activities. On religious NGOs alone, there are Protestant, Catholic, Muslim, and Buddhist organizations operating in the region. Thus, it would be interesting to explore the forming of swift trust in such a complex and dynamic environment.

We decide to perform a survey rather than a case study, the common approach in most humanitarian logistics studies for a number of reasons and challenges. A case study involving multiple languages and levels of interpretation is often difficult to undertake. At the same time, a standardized questionnaire with well-designed constructs and larger sample size would make the study results more objective and empirical verification easier in the future.

The questionnaire for this survey was developed through a rigorous process of screening the survey questions for ease of response from the field. In addition to the English version, it was translated into the local language, that is, Bahasa Indonesia, for a survey conducted in Indonesia. It would enable us to reach more local or regional NGOs there who are poor in foreign languages, as suggested by Harkness and Schoua-Glusberg (1998). A large-scale survey was initiated in early 2014. Giving the explorative nature of the study and the scant knowledge about the actual work of the humanitarian organizations in the region, we used the humanitarian logistics education centers in Southeast Asian countries to conduct the surveys among the NGOs participating in the training sessions conducted by these centers. A researcher visited these centers with the trainers, distributed the questionnaires, and collected them back on the spot. As most humanitarian organizations are not accustomed to answering public inquiries and surveys, handling questionnaire personally would be more convenient and effective for them. The researcher could also explain and clarify their questions during the process. We managed to collect 90 questionnaires from three training sessions in Singapore (33 responses), Indonesia (24 responses), and the Philippines (33 responses), respectively.

On examining the responses, we find one questionnaire unusable due to the missing of all critical information. Thus, the actual sample size for most analyses is 89. A few questionnaires only answered part of our questions and missed some information, making the sample size for hypothesis testing not exactly equal. So where necessary, we will report the actual sample size used for the results found in each of the tables presented in this chapter. Table 9.1 presents the respondent profiles.

Table 9.1 Organizational affiliation of respondents (n = 80)
Full size table

Among the respondents, nine of them omitted their profile information such as organization type and size. Among the rest (80), the largest group is the NGO sector (including local, regional, international NGOs, as well as multinational organizations like the UN agencies) (50%), followed by the public sector (28% from government agencies and related associations), academia (14%), and finally the private sector (both corporations and private foundations, 9%). It largely fits with the overall humanitarian landscape in the region where the NGOs and governments are major players.

Measures

The measures for the constructs such as swift trust and coordination are drawn from the extant literature with some adjustments according to the context of emergency relief operations. On swift trust, we adopt three items from Robert et al. (2009):

  • My colleagues who might interact with them would probably consider them trustworthy.

  • Given their track record, I see no reason to doubt their competence and preparation for the task.

  • If I were working with them on a specific task, I believe I can rely on them not to cause me trouble by careless work.

The degree of coordination is measured by the frequency and openness of information sharing, readiness to maintain a positive relationship, proactive in information sharing, and accessibility according to Hossain and Uddin (2012), and Rampersad et al. (2010). The last construct, coordination effectiveness, is developed based on Hossain and Uddin (2012) and Rampersad et al. (2010) with some adjustments to the humanitarian context with the following five items:

  • The relief activities of our organization are well coordinated with the new partner.

  • The relief activities of our organization with the new partner are well coordinated within the humanitarian network we belong to.

  • There exists an effective centralized coordination body in our humanitarian network.

  • The centralized coordination body can process all the information from the network.

  • The centralized coordination body can coordinate well with the new partner.

Results

Based on the survey results, we first test hypotheses H1–H4, with answers from the first survey question, “based on your experience, when you work with organizations or persons you did not know previously, to what extent do the following factors induce your trust in your unknown partners?” Factor analysis through principal components analysis managed to derive factors from the seven items used in the survey. The result is presented in Table 9.2 with four loaded factors. As these factors account for 81% of the observed variance in the data and the Kaiser-Meyer-Olkin (KMO) measure is 0.64, the exploratory factor analysis is deemed valid.

Table 9.2 Factor analysis on factors that induces trust (n = 88)
Full size table

The three items in the first factor, “the person is from an organization I know,” “the person is introduced by someone I know,” and “I have some good friends in the organization,” are used to measure the variable for H1, “introduced by trusted person.” Reliability analysis also supports the grouping with the Cronbach’s alpha at 0.76. The two items in the second factor are conceptually inconsistent as the first “competency by background or experience” is quite different from the second “sharing same values” and reliability analysis does not support the grouping either with Cronbach’s alpha at 0.68, below the threshold for item consistency. Moreover, both factors three and four contain single items, and we thus treat the rest four items individually.

For the testing of H2, H3a, H3b, and H4, the single items in the questionnaire are therefore used. The items “the organization follows similar rules or procedures in humanitarian operations as mine,” “we share the same values with the organization,” “we share the same religious beliefs with the organization or the person,” and “given his/her background and experience, the person is competent in humanitarian operations” are employed for H2, H3a, H3b, and H4, respectively.

We then conduct t-tests to test the first four hypotheses as we have single variable for each hypothesis. The result is presented in Table 9.3, which shows that H1, H2, H3a, and H4 are supported while H3b is not. H1, H2, H3a, H4 are all strongly supported when compared with the null hypothesis (mean = 3). On the other hand, H3b is strongly rejected and religious belief has been viewed as having no impact on swift trust generation.

Table 9.3 t-Test results for H1–H4 for null hypothesis (mean = 3) (n = 89)
Full size table

To test H5, we use the question “based on your experience, when you work with organizations or persons you did not know previously, to what extent do the following factors lower your trust in your unknown partners?” Two items “potential competition for funding” and “potential clash due to differences in value” are used to measure the variables for H5a and H5b, respectively. t-Tests are similarly employed, and the results in Table 9.4 show that H5a is rejected but H5b is supported.

Table 9.4 t-Test results for H5 for null hypothesis (mean = 3) (n = 88)
Full size table

For the testing of H6 and H7 on coordination activities, a factor analysis is conducted by principal components analysis to derive the coordination variables from the multiple items reported in the survey. The result is presented in Table 9.5 with three loaded factors. As these factors account for 77% of the observed variance in the data and the KMO measure is 0.75, the explanatory factor analysis is deemed valid.

Table 9.5 Factor analysis on coordination activities (n = 84)
Full size table

The five items in the first factor, coordination with the new partner, coordination in the network, effective centralized coordinator, centralized coordinator with information, and good coordination with the new partner by the central body, are measures for coordination effectiveness. The second factor includes the openness, regularity, and proactivity of information sharing, which measures the different aspects of information sharing, an aspect of coordination. The third factor includes approachable and active assistance for the new partners, measuring the active assistance in coordination. Reliability analysis supports the grouping of variables with Cronbach’s alpha at 0.92 for coordination effectiveness, 0.85 for information sharing, and 0.70 for active assistance, respectively, which supports their internal consistencies.

After the factor analysis, we average the item scores for the three factors and derive the value of all coordination variables, coordination effectiveness, information sharing, and active assistance, respectively. H6 is thus further divided into two sub-hypotheses:

H6a: The higher the swift trust among the coordinating partners, the higher the degree of their coordination activities in information sharing.

H6b: The higher the swift trust among the coordinating partners, the higher the degree of their coordination activities in active assistance.

Similarly, hypothesis H7 is divided into two sub-hypotheses:

H7a: The higher the degree of the coordination activities in information sharing, the better the coordination effectiveness.

H7b: The higher the degree of the coordination activities in active assistance, the better the coordination effectiveness.

Four linear regressions are conducted to test the four hypotheses with organizational type (public, private, NGOs, and others) as the control variable. The results are reported in Tables 9.6 and 9.7.

Table 9.6 Linear regression on H6 (n = 86)
Full size table

Tables 9.6 and 9.7 show that H6a, H7a, and H7b are supported by regression analysis but not H6b, and control variable organization type has no impact on our dependent variables.

Table 9.7 Linear regression on H7 (n = 86)
Full size table

Discussion

With a reasonable sample size, we are able to conduct more rigorous empirical analysis on the forming and impact of swift trust in the field of humanitarian operations. In general, most of our hypotheses are well supported by the empirical data. Introduced by a trusted person or organization can lead to higher trust (H1), similarity in rules and procedures can generate trust (H2), sharing similar organizational value can generate trust (H3a), and a competent partner is more likely to be trusted (H4). However, there are also surprises. Contrast to our expectation (H3b), religious belief or values have no impact on the forming of swift trust. Actually, most respondents strongly agree that religion should not have any impact on their professional behavior. In another similar question on the barrier to the forming of swift trust, most respondents also strongly disagree that religious differences should be a source of organizational conflict. It shows that humanitarian workers are largely professional in their views on religion. In Southeast Asia, a region with many religions and occasional religious conflicts, the humanitarian workers in general are able to put religion into their private domain and maintain a professionally neutral position when dealing with organizations and folks from different religious backgrounds.

Our test on H5a and H5b also shows the differences between principles and monetary benefits. While organizational value differences is an important barrier to the forming of swift trust (H5b), potential conflict in fund raising (H5a) is not. Humanitarian workers are not so concerned about the competition for funding among the humanitarian organizations, but focus more on the ground operations which demand high trust among the different organizations. Here again, we note the professionalism of the humanitarian staff.

Moving from the forming of swift trust to its impact, H6 and H7 present a mixed picture. While swift trust is found to be inductive to coordination in information sharing (H6a), its linkage to active assistance in coordination is not supported by the data (H6b). It shows the limitation of swift trust in the field. While humanitarian organizations are willing to trust unfamiliar partners quickly for the groundwork, their coordination is still limited in scope. Some swift trust is good enough for low cost coordination like information sharing, but is insufficient for the more costly forms of coordination such as active assistance. Especially due to the chaotic nature on the ground for emergency relief operations, it is difficult for an organization to spend additional time and effort on unfamiliar partners when there are tons of tasks already in the activity pipeline. Permanent and long-term trust and close relationships from past interactions may be necessary for more active assistance in the coordination. Moreover, both H7a and H7b are supported, showing that coordination in either information sharing or active assistance can enhance the effectiveness of coordination.

Conclusion

The humanitarian aid supply network in disaster relief operations is a typical HFN where members come from organizations with different backgrounds and organizational culture. Nurturing swift trust in such a group is critical for the coordination in the network and thus its effectiveness in the relief operation, particularly so when it concerns rapid response situations. This study investigates the antecedents of swift trust and develops an empirically testable framework linking trust, coordination activity, and performance. We conducted a survey among the humanitarian organizations in Southeast Asia, yielding 89 usable responses. The results show that most hypotheses on swift trust generation are supported, with some interesting exceptions as well. However, being the first empirical investigation without a large sample size, the usual caveat applies and thus limits our conclusion and generalizations.

Our findings highlight the importance of swift trust in humanitarian operations and identifies several means to enhancing the trust. NGOs and governments should use their means effectively to improve swift trust among the humanitarian players. For example, third-party certification and personal competency are inductive to swift trust. Organizing more external training activities would be beneficial for humanitarian workers in both network building, communication, coordination understanding, and competency enhancement.

Clearly, as with all empirical studies, the richness of the sample and the size of the sample do help in providing more accurate and deeper insights into the phenomenon of study. In this chapter, our small sample size naturally calls for a guarded conclusion of our results.

Future studies could increase the sample pool and examine the empirical results more rigorously, and test the five antecedent conditions for swift trust with better proxies, drawn from other domains of study such as Fischer (2013) who has dealt with the issues of trust, communication, and past collaboration. The connection between swift trust and coordination effectiveness can be explored in depth, following the footsteps of Gulati and Singh (1998) and other management scholars. It is also interesting to note that religious belief has no statistical impact on swift trust generation. It would be interesting to explore this anomaly more carefully at the country level and explore the means to achieving such a harmony which is one hallmark of the professionalism of the NGOs. We defer this to another study.