Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan

Technology in Society xxx (2014) 1e11

Contents lists available at ScienceDirect

Technology in Society journal homepage: www.elsevier.com/locate/techsoc

Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan Kent Marett a, *, Allison W. Pearson a, Rodney A. Pearson a, Erich Bergiel b a b

Department of Management & Information Systems, College of Business, Mississippi State University, Box 9581, MS 39762, USA Department of Management, Richards College of Business, University of West Georgia, Carrollton, GA 30118, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 May 2014 Received in revised form 19 November 2014 Accepted 26 November 2014 Available online xxxx

Mobile phone adoption and use are common-place in the western world, yet still are associated with risks of loss of privacy and information security. However, in high-risk cultures and countries, such as those at war or threatened by terrorism, mobile phone adoption and benefits of use may be perceived quite differently. In this study, we use ecommerce and adoption theories to build a model of trust and risk as predictors of mobile use benefits in a sample of current mobile users in southern Afghanistan. The responses collected from a survey of over seven thousand Afghani citizens were used to test the research model. The results suggest that despite the potential danger, the mobile device owners who were surveyed perceived the benefits derived from use as being worthwhile. The results are discussed with implications for managers and practitioners provided. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Mobile use Risk Trust Perceived benefits Afghanistan

1. Introduction The vast majority of research on business communication, including research investigating adoption and use of communicative technology, has been conducted within highly-developed, stable, capitalist societies. In fact, most research assumes that individuals have free speech rights, basic safety, and protection of their lives and business provided by the government. As other researchers have pointed out [1,2], we know little of the day-to-day communicative practices of individuals located in less developed, unstable, high-risk societies, even individuals in those regions who are members of multinational firms. The Fund for Peace identifies 67 countries around the globe that fall into a high-risk context based on groups of citizens who are vengeance e seeking, struggling economic

* Corresponding author. þ1 662 325 7001. E-mail addresses: [email protected] (K. Marett), [email protected] (A.W. Pearson), rpearson@business. msstate.edu (R.A. Pearson), [email protected] (E. Bergiel).

development, poverty, deteriorating government, and violations of human rights. In this study, we explore the fundamental risks and dangers associated with mobile technology usage in a high-risk environment, fraught with war, terrorism, physical violence, and extreme poverty. Given that internal communication between members of multinational corporations come under the same strains of violence, terrorist attacks, and the inability of government to protect its infrastructure as the rest of the citizenry located in those areas, how does our current understanding of communicative technology usage apply to high risk environments? Earlier work in mobile networking has called for increased attention toward cultural differences and perceptions of mobile services [3]. One of the areas in which cultural differences may manifest is with perceptions of privacy and safety. In many parts of the world, mobile users may take for granted that cell networks are secure and that providers will protect one's privacy, while in other regions a lapse in privacy may affect one's personal safety. For example, during focus group interviews

http://dx.doi.org/10.1016/j.techsoc.2014.11.002 0160-791X/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

2

K. Marett et al. / Technology in Society xxx (2014) 1e11

conducted with a wide range of mobile users in Scandanavia, North America, and Eastern Asia, vulnerabilities were described as one's losing a phone and losing touch with others, and privacy issues took the form of discussing sensitive topics in public places around random bystanders [4]. As past events have shown, mobile users in the Western world may have a false sense of security, and that in more unstable countries and contexts, risks of mobile use may greatly differ. In this study, we focus on mobile users who have reason to believe their cellular service provider makes their personal information and privacy vulnerable. To our knowledge, the literature on security and privacy in a mobile environment primarily focuses on the endpoints of the communication. However, there is every reason to believe that the vulnerabilities to a user may lie within the network. Mobile users frequently roam through multiple cells during a connection to the network, making them susceptible to malicious or compromised cell domains with the potential of engaging in information theft and denial of service, among other threats [5]. Researchers have also documented the possibility of remotely draining the battery power of others' mobile devices through “ping-of-death” vulnerabilities in cellular networks [6]. Reports indicate that mobile users have been unknowingly monitored through the use of location-based services implemented on their devices [7]. However, a more likely threat to privacy lies within the network providers themselves. Social engineering attacks on service providers have exploited holes in security policies protecting mobile user accounts in North America and Europe [8,9]. The providers themselves may be intentionally responsible for security breaches, including those originating from employees within the company [10]. When trust in mobile service providers is breached, mobile use and benefits of mobile use may decline. These effects may be even more pronounced in high risk contexts. 2. Literature and theory review The widespread adoption of mobile devices is said to be “volcanic” in developing countries [11]. However, in less affluent countries stricken with political instability, war, and terrorism, mobile technologies may have different implications toward usage. Conflicting pressures from countries who want to provide information access and external threats, such as terrorist groups, who seek to reduce or control information access, result in difficult initial decisions to adopt [12] for citizen mobile users. Nonetheless, the widespread adoption and use of mobile phones can be linked to societal change including economic growth, improved medical care, and reductions in poverty [13,14]. In developing nations and war-torn nations with limited communications infrastructure, there is a powerful need and desire for access to information by the citizenry, in spite of the risks (such as terrorism) associated with use. The perceived benefits of use of mobile technologies for access to people and information are likely strong drivers of user behavior. Yet, very little is known about the impact of external, violent threats with regard to

information access on mobile usage, even though it is widely acknowledged that such violence hinders entire economies [12,15]. Indeed, the adoption of IT in emerging and dynamic domains is a relatively unexplored area of research [16]. The purpose of our study is to explore the benefits and risks of mobile phone use, in a high-risk context, based on a sample of mobile users in Southern Afghanistan e a region that is unstable and with ongoing terrorist threats and violence. Afghanistan is one current example of a high risk context, as it is considered to have a government, the Taliban, that supports state-sponsored terrorist activities [17]. Under pressure and threats of violence from the Taliban, mobile service providers shut down mobile towers at night, often from 5:00pm e 8:00am, and in some rare cases, the signal may be available only 4 h per day [18]. The Taliban forces mobile providers to shut down towers by threatening physical violence and property destruction on both mobile employees and mobile providers who do not comply with their demands. A manager for one of the major mobile providers reported to the New York Times that unless the towers are shut down at night, the Taliban promises that “employees will be abducted, killed, and the towers will be burned,” a threat that was realized when three cell towers in the Kapisa Province were bombed and destroyed [18]. The control of mobile signal access is reported to send a strong psychological message to Afghani citizens that the Taliban can have direct control over their future. Further, the mobile shut down also creates doubt about whether or not the Afghanistan government can protect citizens, further increasing the fear and anxiety of mobile users. In the next section, we build a theory base upon which to explore why, in spite of such high risk, mobile users in high risk contexts will still see benefits in use of mobile technologies. 2.1. Theory of risk and use of technology Researchers in the technology adoption literature have pointed out, perceived risk of using a technology has often been overlooked or minimized in lieu of focusing on the benefits stemming from adoption e.g., [19,20]. These criticisms have led to the conclusion that the core model of adoption theory “has limited usefulness in the constantly evolving IT adoption context” [[21], p. 212]; however, they may be adapted to understand mobile use in high risk contexts. For consumers of mobile services, perceived risk is defined as the “uncertainty regarding possible negative consequences of using a product or service” [[19], p. 453]. Researchers have since theorized the integration of technology adoption models and risk theories to explain the likelihood of on-line adoption, mainly focusing on purchase intent of consumers. Perceived risk is most prevalent during the decision to adopt and use IT when feelings of uncertainty, discomfort, anxiety, and conflict exist within the user [19]. Risks perceived by technology users are often centered around possible task inefficiencies coupled with the risk of unsecured communication, potential loss of private information, and possible financial losses [22]. The basic assumption in the e-commerce literature is that system

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

K. Marett et al. / Technology in Society xxx (2014) 1e11

use and adoption, including purchase decisions, will be negatively impacted by perceived risk of use [23]. As a result of perceived risk, technology adopters consider the consequences of their usage to be of such significance that the technology providers must address the risk directly. For instance, customers who disclose personal information to companies run the risk that the information will not be used in an appropriate manner, so companies are often required to notify customers of their information use practices [24]. However, mobile users put more than their own personal information at risk; a user's list of contacts and related identifying information is often stored in centralized data repositories managed entirely by the carrier, and users' voice and data communication is transmitted across the cellular network maintained by the carrier. Thus, it should be expected that the perceptions of the mobile service provider will influence the amount of risk perceived by mobile users. Perceived risk is typically associated with the expectation of loss [12]. There are, however, a variety of types of risks e some of which are more relevant in an unstable or developing country than others. Financial, product and information privacy risks are the predominant risks found in e-commerce research [22]. Product or performance risk is defined as the probability that the product purchased on-line will fail or not function as desired [19]. Performance risks associated specifically with mobile use also include danger, antisocial behaviors and distraction [25]. Financial risks include the risk of value e overpaying for an underperforming item. Mobile technology use can result in financial risk for those individuals in developing nations who lack stable or sufficient income to provide for their families. Additionally, in hostile regions, physical threats associated with technology use or information collected through technology use are paramount. Psychological threats are probable, as well as lack of privacy, reduced security, and possible social ostracism for behavior that is unconventional with regard to cultural norms. In total, these risks of adopting mobile devices are sizeable, and quite different from the risks hypothesized in E-Commerce theory with on-line purchasing. Yet, mobile phone use does exist under these risky conditions. When explaining why individuals will adopt technologies that may exist in a risky context, the relationship between two important variables, risk and trust, must be examined. The relationship between risk and trust has been hypothesized in the online environment with online use/adoption theories [26,27], as well as with interorganizational data exchanges [28]. Previous ecommerce research hypothesizes that trust, perceived risk and perceived benefits of technology influence purchase intentions, and that trust influences purchase intentions by the mediating relationship with risk [22]. However, these relationships have not been examined with mobile phone technologies and have not been examined in unusually high risk contexts. With regard to mobile phone technologies, a key question to explore is “why risk and trust beliefs arise and how IT artifacts can be utilized according to the perceived sources of these beliefs” [[26], p. 281].

3

3. Hypotheses 3.1. Risk and benefits of use Much as potential mobile users likely consider the risks involved when deciding to use the technology, the perceived benefits are likely appraised as well. Previous literature suggests that risk is negatively associated with perceived benefits; in fact, a frequent finding is that the greater the perceived benefits of mobile technology are, the lower (or more acceptable) the perceived risks are, although this relationship is void if the risks are overly severe [13]. We expect that this relationship is reciprocal, and that the greater the perceived risk of mobile use is, the lower the expectations of the perceived benefits will be. Indeed, earlier work provides evidence that perceived risk of mobile use reduces the intention to purchase a phone in the first place [29], even before benefits can be realized. Other studies provide evidence of the negative relationship between perceived risks and benefits of technology usage. Featherman and Pavlou [19] propose that the performance expectancy of a system under scrutiny for adoption is negatively affected by perceived risk. This may be due to perceptions of uncertainty and potential loss, both of which are related to perceived risk. Uncertainty is commonly a concern with wireless networks and technologies, which are often beyond the control of the individual user, and thoughts of potential losses that may transpire during a transaction occur to the mobile user as a result of uncertainty, thus affecting expectations of the usefulness of the mobile service [23]. One could argue a similar expectation for mobile phone use e that is, increased risks associated with use will lead to lower use and/or lower perceived benefits of use. H1: Perceived risks of mobile phone use are negatively related to the perceived benefits of use. However, our goal is to explore those who have already adopted in spite of the often very high risks. What can lead individuals to use mobile devices in spite of the risks? 3.2. Trust and risk Trust is frequently defined as “a willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that party” [[30], p. 712]. Trust is essential in relationships, such as online transactions, where significant risks arise based on uncertainty, interdependence, and possible opportunism [31]. Trust is therefore necessary to reduce uncertainty and risk [32], and it has frequently been used in the e-commerce literature to explain how consumers can overcome risks in order to participate in online transactions [23,33]. In fact, trust is deemed truly operable only in the presence of risk [12]. Others have purported that trust leads to risk taking. Social exchange theory [34] suggests that when uncertainty and risk exist in an exchange relationship, trust has the effect of reducing uncertainty and risk, so that utility can be fully realized [35]. The ability of trust to reduce risks

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

4

K. Marett et al. / Technology in Society xxx (2014) 1e11

with regard to technology usage has received robust empirical support within the literature [e.g., [31,36e39]]. With regard to cell phone use in high risk environments, we then propose that trust in the mobile service provider and the accompanying network will reduce perceived risk of use, and subsequently may lead to increased perceived benefits of cell phone use. It is only with high levels of trust that a citizen may be willing to overcome risks that they are aware of, but not able to control, such as terrorists threats. We can express this relationship as risk will mediate the general relationship of trust and perceived benefits of use. H2: Trust in the mobile phone provider will be negatively related to Perceived Risk of mobile phone use. Therefore, Perceived Risk will fully mediate the relationship of Trust and Perceived Benefits of Use of mobile phone users in high-risk contexts. However, there are different forms of trust that may impact this general relationship in unique ways. In models of organizational trust [30], the characteristics, actions, behaviors, and competencies are sources of trust. Characteristics of organizational trust can be based in the ability, benevolence, and integrity of the organization. When operationalizing trust, Lee and Rao [12] decomposed trust into the specifics of belief in good intentions, belief in domain competence, and belief in online competence. Other researchers have suggested that trust and distrust are distinct dimensions and should be measured separately [26]. For this study, we conceptualize 3 separate forms of trusts in high-risk contexts as (1) trust in the competence of the mobile provider, (2) trust in the general intentions of the provider, and (3) distrust in the provider's actions. 3.2.1. Trust in competence The organization trust literature purports that organizations demonstrating abilities and competency in core technical areas results in trust from customers and users. Ecommerce researchers have operationalized the ability dimension of trust as “belief in competence” [12], and defined it as the user's general belief of the capability and quality of on-line services provided. Service provider domain competence is linked to the user's intention to use a web-based service. Other researchers have operationalized the ability form of trust in ecommerce sites as accuracy and completeness of website information [22]. It has been recognized that content providers, in an effort to gain the trust of mobile users, should attempt to reinforce authenticity and provide structural reassurances of competency in order to reduce perceptions of risk [40], and we expect the same advice should hold true for service providers. A high level of trust in a company's competency reduces a fear of the unknown, making the user feel more in control and at less risk [41]. We operationalize Trust in the Competence of the mobile service provider by assessing the user's perception of how competent the provider is in delivering customer service, signal coverage, reliability and other aspects of providing basic service. As a form of trust, we hypothesize that Trust in Competence of the provider will reduce the perceived risk of mobile use. H2a: Trust in Competence of the mobile provider will be negatively related to Perceived Risk of mobile phone use, such that Perceived Risk will fully mediate the relationship

of Trust in Competence and the Perceived Benefits of Use of mobile phone users in high-risk contexts. 3.2.2. Trust in the institution Institution-based trust is a sense of confidence in the provider's expected behavior and confidence in the provider's goodwill [27,42]. Institution-based trust accrues when the company acts as it should e it respects societal values, operates in conjunction with societal norms of economic transactions, and in short, fulfills its expectations. This dimension of trust has been operationalized as the positive reputation of the selling party where it “has been considered a key factor for reducing risk and creating trust” because it represents “the degree of esteem in which consumers hold a selling party” [[22], p. 551]. Trust in the provider institution relates to the trustor's first-hand knowledge of how the company operates [27]. Based on repeated interactions with the mobile provider, trust develops gradually over time with each experience, and fully developing once mobile users have adopted and used mobile technologies and develop an interaction history with the provider. Trust in the institution has been hypothesized as reducing risk perceptions of users, which in turn produces greater use [13]. Other researchers have proposed that trust in the provider institution reduces the perceived risk of privacy issues [12]. Interviews with mobile users provide evidence of trust in the service provider reducing perceived risk, with users more likely to utilize mobile payment services using the networks of wellestablished and reliable providers [43]. As a form of trust, we propose that Trust in the Institution of the provider will reduce the perceived risk of mobile use. H2b: Trust in the Institution of the mobile provider will be negatively related to Perceived Risk of mobile phone use, such that Perceived Risk will fully mediate the relationship of Trust in the Institution and the Perceived Benefits of Use of mobile phone users in high-risk contexts. 3.2.3. Distrust in the institution Distrust itself can be defined as “the expectation of assured negative behavior by the other party and the apprehension that the other party will harm the interests of the party that created the trust” [[44], p. 179]. Distrust is a distinct dimension of trust, and it is not considered the low end of the trust continuum. Instead, it can be argued that distrust exists even when trust is non-existent [26]. Further, distrust resulting from negative events is often more salient to the user than positive events that could lead to trust [44]. Similar to the attribution of trust one can place on an institution, distrust in an institution can occur when one confidently feels conditions are conducive to situational failure in a risky transaction with said institution [45]. While trust leads to an assumption that expected and acceptable behavior will follow, and hence risks are reduced, distrust leads to an increased emphasis on likely risks and negative outcomes. Distrust leads to avoidance behavior in order to minimize or eliminate risks, while at the same time discourage behaviors directed at positive outcomes, such as sharing information, delegating responsibility, and ultimately a willingness to conduct

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

K. Marett et al. / Technology in Society xxx (2014) 1e11

transactions [46]. Therefore, we propose that Distrust in the provider will increase the perceived risk of mobile use. H2c: Distrust in the Institution of the mobile provider will be positively related to Perceived Risk of mobile phone use, such that Perceived Risk will fully mediate the relationship of Distrust and the Perceived Benefits of Use of mobile phone users in high-risk contexts. Our hypothesized relationships regarding trust are detailed as follows and are shown in Fig. 1 below. 4. Method The researchers drew from a secondary data source. Data collection was conducted via face-to-face interviews by the Human Terrain System Social Science Research and Analysis team between April 19 and May 3, 2011 in 19 districts in the Afghan Helmand and Nimroz provinces (see Fig. 2). A total of 7022 individuals participated; of these, 3033 (43.2%) owned a mobile phone, either personally or through family members. Following the guidance of Mayer and colleagues [30], the effects of risk and trust should only be evaluated “in terms of actual behavior, not willingness to engage in behavior” (p. 729), in order to capture the actual willingness to be vulnerable. Therefore, responses from the 3033 mobile phone owners were used in this study. Survey respondents were primarily male (64.5%) and relatively young (69.6% between the ages of 20 and 39). Fifty-four percent of the respondents have no formal education; only 1.9% have formal education of more than 12 years of school. For comparison purposes, worldwide estimates indicate 24.7% of the population aged 25 and over have no formal education, and 14.3% have education beyond secondary schools [47]. Over half of the survey respondents have problems reading a letter (55.8%), writing a letter (56.5%), or reading a book (55.7%).

5

Thirty-three percent of respondents work full-time, not counting 11.4% who work as full-time farmers. 16.7% work part-time. 33.7% are housewives not working outside the home. The vast majority of respondents have limited access to electricity, with 92.9% having access fewer than six days per week. Seventeen percent of respondents have no access to electricity. Not surprisingly, given limited access to electricity at home, 22.9% of respondents charge their devices somewhere other than home or work, using options such as charging by battery (5.8%), via a street vendor (5.2%), in a vehicle (3.6%), solar charging (2.4%), or in a mobile phone shop (2.3%). Forty percent of respondents have a personal monthly income of less than $250 U.S, and 81.9% have income of under $700 per month U.S. While Afghan mobile phone costs are lower than U.S. costs (60.5% of respondents pay less than $5.25 U.S. per month), respondents did report having problems paying for other essential living expenses because of their mobile phone expenses, with 7.0% reporting problems paying for food. 4.1. Measures The five constructs of the study were assessed with Likert-scale questions, using a four point response set where 1 ¼ strongly disagree and 4 ¼ strongly agree. Previously developed and validated scales based on the ecommerce literature or the technology usage literature that were primarily developed in the western world and in stable, corporate settings, did not seem applicable in the high risk context under study. Therefore, the measurement items for each construct were designed uniquely for this study. All items are shown in the Appendix. Trust in Competence of the provider was measured with four items regarding signal coverage, price, customer service, and signal reliability. Trust in the Institution of the provider was

Fig. 1. Perceived benefits of mobile use in a high risk context.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

6

K. Marett et al. / Technology in Society xxx (2014) 1e11

Fig. 2. Location of survey respondents.

measured with four items designed to tap into the degree to which the provider engenders confidence based respect for societal values and other normative economic expectations. Items include “Afghan mobile providers respect the values of Islam” and “Afghan mobile providers are good companies to work for”. Distrust in the Provider was measured with 3 items designed to measure the level of perceived intentional harmful or misleading behaviors of the provider. A sample item is “Afghan mobile providers allow others to listen in on private conversations”. The mediator construct, Perceived Risks, was measured with 3 items, including such items as “Improved mobile services will help the Taliban/AGE to better plan out attacks”. The

dependent variable, Perceived Personal Benefits, was measured with 3 items, and included the sample item “More reliable mobile phone service will improve my overall quality of life”. 5. Results Structural Equation Modeling (SEM) using partial least squares (PLS) was the analysis technique selected to test the hypothesized relationships. PLS allows for simultaneous estimation of all hypothesized relationships, including error estimation. The PLS approach includes two separate assessments, the measurement model, which

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

K. Marett et al. / Technology in Society xxx (2014) 1e11 Table 1 Descriptive statistics and inter-construct correlations. Mean (SD) DIS Distrust Trust in institution Trust in competency Perceived risks Perceived benefits

2.42 2.96 2.26 2.05 3.30

(0.7) (0.6) (0.6) (0.8) (0.6)

TI

TC

RISK BENE

0.71 0.06 0.59 0.29 0.17 0.62 0.36 0.30 0.30 0.73 0.11 0.02 0.11 0.15 0.76

Note: Responses range from 1 (strongly disagree) to 4 (strongly agree). Square root of average variance extracted bold in diagonal.

focuses on the reliabilities and validities of the measures used in the research model, and the structural model, which tests the hypotheses predicted by the research model. In terms of assessing the convergent validity for the measures, as shown in the Appendix, each item loaded cleanly on its appropriate measure without cross-loading on another. The discriminant validity of the reflective measures can be evaluated by comparing the square root of the average variance explained by each construct with its correlations with the other constructs in the model. The square root of AVEs were all larger than the crosscorrelations (see Table 1 below), and the correlations among the constructs themselves were all well below 0.90, suggesting discriminant validity. We conducted hypothesis testing of the research model using PLS, applying a bootstrapping technique to estimate the significance of the path coefficients [48]. PLS does not provide goodness-of-fit indices, so the predictive power of the structural model is suggested by the amount of variance explained in the dependent latent variables. In the current study, 23 percent of variance was explained for perceived risk of use, but only 3 percent of variance was explained for perceived benefits of use. Fig. 3 presents the results of the structural model.

7

All four of the hypothesized relationships were found to be statistically significant at the p < 0.01 level. Hypothesis 2A, which predicted trust in the service provider's confidence would negatively influence perceived risk of use, was supported, as was Hypothesis 2B, which predicted the same negative influence from trust in the institution. Hypothesis 2C, which predicted that distrust in the provider would positively influence perceived risk of use, was also supported. Finally, Hypothesis 1, which hypothesized perceived risk of use would negatively influence the perceived benefits gained from cell phone use, was found to be significant but in the opposite direction (b ¼ 0.16). In order to better understand the unexpected result for H1, we conducted a post-hoc analysis of the research model with direct paths added between the endogenous variables and perceived benefits of use. In short, we wanted to explore the possibility that the trust factors mitigate the proposed effects of perceived risk on the dependent variable, perceived benefits of use. This model would be considered a partially mediated model, where the three forms of trust may also have a direct effect on perceived benefits of use. Trust research suggests that such direct effects may exist under some risk conditions. Trust has been suggested to be the mechanism by which risk perceptions can be reduced enough to enable risk taking behavior (Mayer et al., 1995). Adding the direct paths resulted in a higher R2 for perceived benefits of use (0.17 versus the original 0.03). The only variable with a significant direct relationship was trust in good intentions (b ¼ 0.16, p < 0.001). Interestingly, H2B was no longer significant with the direct paths included in the model, suggesting that perceived risk of use does not mediate the influence of trust in the institution. Rather, trusting that the provider has good intentions may overcome the perceived risks involved with cell phone

Fig. 3. Results of hypothesis testing.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

8

K. Marett et al. / Technology in Society xxx (2014) 1e11

use when considering the potential benefits to be gained. The results of the post-hoc analysis exploring the direct effects of trust on perceived benefits of use are shown in Fig. 4. 6. Discussion This study provides clear evidence that cell phone services are thought by individuals to be instruments for the betterment of their lives and for their society as a whole in the high risk context explored in our study. Survey respondents, who are all current mobile users, reported high expectations in terms of the personal benefits available from mobile phone use. Further, although respondents are largely uneducated and unemployed, they reported often giving up necessities in order to afford mobile service and sometimes travel great distances to merely recharge their phones. Despite the personal risks and sacrifices they must make in order to use cell phone service, the respondents have determined that the potential benefits (albeit limited) are worth it. Even within one of the most hostile and dangerous regions of the world, citizens risk their personal safety and convenience in order to seek technological means for connecting themselves with friends, family, and the world beyond their borders. The results here may seem to contradict the trust-riskoutcome framework used for this study, but Mayer and colleagues [30] stipulate that context plays a significant role in risk-taking behavior. Specifically, contextdependent factors such as the personal consequences involved with mobile use and the availability of alternatives to risk takers can override the levels of perceived trust and risk. The results here speak to the importance Afghani

citizens place on mobile connectivity and their desire to achieve Internet access at comparative levels as the rest of the world. Mobile device ownership and usage seem to be persevering in spite of “social location” variables (socioeconomic, demographic, and lifestyle conditions) that have prevented citizens from using such technologies in the past [49], and many Afghanis have found ways to employ mobile technologies under undesirable conditions while placing themselves at personal risk to do so. We believe that the results of this study provide a clear example of the importance of context to the relationships Mayer et al. described. Our findings also confirm the paradox of mobile phones described by Arnold [50] and Prasopoulou and colleagues [51]; while mobile users typically consider their devices to facilitate geographical and social independence, the network architecture and fixed addressing required for their devices also ground them in a codependent relationship with other network nodes. In other words, cell phones may allow users a measure of reassurance through remaining connected to others as they roam, but they also are identifiable to those managing the network and, with recent advances in geolocation applications, other users as well. The results of this study provide empirical evidence that mobile users in high risk environments understand this paradox, yet they actively defy it. The survey respondents reported high distrust in the service providers and perceived high personal risks stemming from mobile usage, indicating they acknowledge potential vulnerabilities to the “network.” However, the perceived need for access to communication and information exchange apparently outweighs those vulnerabilities.

Fig. 4. Post-hoc analysis.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

K. Marett et al. / Technology in Society xxx (2014) 1e11

Finally, in their study of internet diffusion across 143 nations, Robinson and Crenshaw [15] proposed that political conflict (including adverse regime transitions, political violence, terrorism, and war) would negatively impact internet use and ultimately slow societal economic growth. These forms of political conflict often negatively impact IT adoption and use through sabotage of telecommunication infrastructure and by invoking fears of surveillance by either the government or opposition groups. Political conflict often results in losses of individual freedom and constricted information access. While Robinson and Crenshaw found macro-level support for their hypothesis, they also noted that their study was limited due to the inability to “tap the scope or intensity of usage among users” within a conflict-ridden nation. Our study of cell phone users in Afghanistan, which focused on the individual level of analysis, demonstrates that trust in the provider could mitigate some of the risks associated with mobile phone use. Certainly, additional research is needed to increase our understanding of the influence of perceived risk in environments such as the one examined in this study. The use of mobile communication is on the rise in multinational firms, especially with continuing advances in platform-independent information delivery. Despite expectations that the use of mobile technologies will provide for economic growth and improved communication efficiencies within organizations, the potential for such benefits to accrue in the developing world is more limited. Chief among the reasons are infrastructural weaknesses and poor investment in the cellular network, and we believe this study provides ample evidence that an untrustworthy network serves to stymie communication that is taken for granted elsewhere. This problem should not be ignored by people in developed countries, as business organizations and governmental agencies often maintain facilities in underdeveloped regions of the world and, thus, require a secure means for digital information transfer [52]. It should be noted that an unsecure network is not solely a risk for a person residing within a hostile region; because the cellular network is global in scope, individuals who directly interact with the person at risk may also make their own information and communication vulnerable, as well as make “secondary users” who are indirectly affected by the output of the communication vulnerable [53]. This should be an important consideration for managers and employees within multinational firms that partner with companies in developing regions. While the partner could be entirely trustworthy, the security of the network during the “last mile” of communication could be in doubt. This possible vulnerability is the impetus behind implementing virtual private networks and encryption technologies in unsecure international network environments [54]. Mobile users in multinational companies should be aware of the risks involved before establishing communication. We found that in spite of significant perceived risks and distrust in some aspects of the provider, cell phone users in Afghanistan reported moderate levels of trust in their provider and chose to adopt and use cell phones.

9

These users also had to overcome financial risks (7% reported having problems paying for food) and performance risks (22.9% of the users had to charge their phones somewhere other than home or work; some had to use portable generators) associated with mobile use. Perhaps the desire for information, communication, and social connectivity, outweigh the perceived risks of use. Trust in the provider can also further mitigate risks and drive mobile usage. As trust in the provider grows, even in the presence of perceived risks, individuals' use of information technology may continue to increase in high risk contexts. In conflict ridden areas, characterized by war, terrorism, and/or political instability, what can mobile service providers do to increase user trust, when the context itself is one characterized by doubt, distrust, and fear? Fukuyama [35] notes that trust can generalize to a group through normative behaviors demonstrating loyalty, honesty, and dependability. Trusting behaviors carry more influence and develop greater trust when they are evoked by ever larger groups with shared solidarity. Over time, one would expect that continued usage of the mobile network, without incident, will put aside doubts about trusting mobile providers [55]. If trusting behaviors continue unabated, improved trust can eventually extend to group and even societal levels. However, until IT infrastructure is protected from terrorism, distrust in the provider may continue to persist. 7. Conclusion Undoubtedly, IT adoption and use differs across contexts and cultures, and consequently, users in less developed nations may be overlooked in past IT research. Research models of adoption and use were developed in western developed nations and have not been examined thoroughly in developing nations or nations struggling with political conflict, terrorism, or war e those contexts we refer to as high-risk. By examining IT adoption and use models in these high-risk contexts, we may find, as we did in our study, that relationships among constructs of interest are unique. In these high-risk contexts, risks are different, (e.g., the threat of terrorism). Performance and financial risks are also perhaps more extreme. Trust in the mobile provider may play a far more vital role in driving use behaviors, when there are no or few competitors. As researchers, if we continue to overlook these unique, hard-to-study, high-risk contexts, we risk ignoring the IT needs, desires, and uses of those who may not currently have the advantages of a developed, stable nation. While our study has limitations, it provides a preliminary and exploratory view of mobile users in a high risk context and explores their perceptions of trust in the mobile providers and perceived risk of use. This unusual context and findings provide a rare look into the factors regarding mobile use that may impact information access, business viability, economic and societal success when we apply theories of adoption and use of technology outside of our relatively secure environments.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

10

K. Marett et al. / Technology in Society xxx (2014) 1e11

Appendix. Survey instrument.

Factor Distrust in the institution (Cronbach's a ¼ 0.65) Afghan mobile phone providers misrepresent their signal coverage Afghan mobile phone providers charge more than they advertise Afghan mobile phone providers allow others to listen in on private conversations Trust in the institution (Cronbach's a ¼ 0.69) Afghan mobile phone providers provide an important service Afghan mobile phone providers are good companies to work for Afghan mobile phone providers charge prices that are fair Afghan mobile phone providers respect the values of Islam Trust in competency (Cronbach's a ¼ 0.86) How would you rate the quality of the following services provided by Afghan mobile service providers? Signal Coverage Price Customer Service Signal Reliability Perceived risks (Cronbach's a ¼ 0.65) Improved mobile phone service means that youth will be exposed to inappropriate influences The Afghan government listens in on mobile telephone calls in my area Improved mobile phone service will help the Taliban/AGE to better plan out attacks Perceived benefits (Cronbach's a ¼ 0.86) Wider coverage of mobile phone service makes me feel safer More expansive mobile phone service will help businesses to serve customers like me More reliable mobile phone service will improve my overall quality of life

Loading 0.76 0.83 0.47

0.74 0.70 0.52 0.70

0.89 0.84 0.86 0.57 0.83

0.56 0.80

0.61 0.81

0.83

References [1] Soltani E, Wilkinson A. The Razor's edge: managing MNC affiliates in Iran. J World Bus 2011;46(4):462e75. [2] Goodman M. Corporate communication in times of stress. Corp Commun An Int J 2011;16(3). [3] Ngai EWT, Gunasekaran A. A review of mobile commerce research and applications. Decis Support Syst 2007;43(1):3e15. [4] Jarvenpaa S, Lang K, Tuunainen VK. Friend or foe? the ambivalent relationship between mobile technology and its users. In: Sorensen C, Yoo Y, Lyytinen K, DeGross J, editors. Designing ubiquitous information environments: socio-technical issues and challenges. New York, NY: Springer; 2005. p. 29e42. [5] Ghosh A, Swaminatha T. Software security and privacy risks in mobile e-commerce. Commun ACM 2001;44(2):51e7. [6] Racic R, Ma D, Chen H. Exploiting MMS vulnerabilities to stealthily exhaust mobile phone's battery. In: Second International Conference on Security and Privacy in Communications Networks; 2006. Baltimore, MD, USA. [7] Junglas I, Johnson N, Spitzmüller C. Personality traits and concern for privacy: an empirical study in the context of location-based services. Eur J Inform Syst 2008;17(4):387e402. [8] Goggin G. Cell phone culture: mobile technology in everyday life. New York, NY: Routledge; 2006.

[9] Robinson J. Now phone-hacking scandal: news Corp's 'rogue reporter' defence unravels. Available from:. 2011. http://www. guardian.co.uk/media/2011/jan/17/phone-hacking-news-of-theworld?CMP¼twt_fd [Last accessed 11 November 2014]. [10] Willison R, Siponen M. Overcoming the insider: reducing employee computer crime through situational crime prevention. Commun ACM 2009;52(9):133e7. [11] Abu-Shanab E, Abu-Baker A. Using and buying mobile phones in jordan: implications for future research and the development of new methodology. Technol Soc 2014;38:103e10. [12] Lee J, Rao HR. Perceived risks, counter-beliefs, and intentions to use anti-/counter-terrorism websites: an exploratory study of government-citizen online interactions in a turbulent environment. Decis Support Syst 2007;43:1431e49. [13] van Kleff E, Fischer ARH, Khan M, Frewer L. Risk and benefit perceptions of mobile phone and base station technology in Bangladesh. Risk Anal 2010;30(6):1002e14. [14] Black J, Koch F, Sonenberg L, Scheepers R, Khandoker A, Charry E, et al. Mobile solutions for front-line health workers in developing countries. In: Proceedings of the 11th International Conference on e-Health Networking, Applications, and Services; 2009. Sydney, Australia. [15] Robinson K, Crenshaw E. Reevaluating the global digital divide: socio-demographic and conflict barriers to the Internet revolution. Sociol Inq 2010;80(1):34e62. [16] Wells J, Campbell D, Valacich J, Featherman M. The effect of perceived novelty on the adoption of information technology innovations: a risk/reward perspective. Decis Sci 2010;41(4):813e43. [17] Branscomb L. Protecting civil society from terrorism: the search for a sustainable strategy. Technol Soc 2004;26:271e85. [18] Rubin AJ. Taliban using modern means to add to sway, in New York Times. 2011. New York, NY, USA. [19] Featherman M, Pavlou P. Predicting e-services adoption: a perceived risk facets perspective. Int J Human Comp Stud 2003; 59(4):451e74. [20] Im I, Kim Y, Han H. The effects of perceived risk and technology type on users' acceptance of technologies. Inform Manag 2008;45:1e9. [21] Benbasat I, Barki H. Quo vadis TAM? J AIS 2007;8(4):211e8. [22] Kim DJ, Ferrin D, Rao HR. A trust-based consumer decision-making model in electronic commerce: the role of trust, perceived risk, and their antecedents. Decis Support Syst 2008;44(2):544e64. [23] Lou X, Li H, Zhang J, Shim JP. Examining multi-dimensional trust and multi-facted risk in initial acceptance of emerging technologies: an empirical study of mobile banking services. Decis Support Syst 2010; 49:222e34. [24] Culnan M, Armstrong P. Information privacy concerns, procedural fairness, and impersonal trust: an empirical investigation. Organ Sci 1999;10(1):104e15. [25] Middleton C, Cukier W. Is mobile email functional or dysfunctional? Two perspectives on mobile email usage. Eur J Inform Syst 2006; 15(3):252e60. [26] Gefen D, Benbasat I, Pavlou P. A research agenda for trust in online environments. J Manag Inform Syst 2008;24(4):275e86. [27] Gefen D, Karahanna E, Straub D. Trust and TAM in online shopping: an integrated model. MIS Q 2003;27(1):51e90. [28] Nicolaou A, McKnight H. Perceived information quality in data exchanges: effects on risk, trust, and intention to use. Inform Syst Res 2006;17(4):332e51. [29] Teng W, Lu H-P. Consumer adoption of PDA phones in Taiwan. Int J Mob Commun 2010;8(1):1e20. [30] Mayer R, Davis J, Schoorman D. An integrative model of organizational trust. Acad Manag Rev 1995;20(3):709e34. [31] Pavlou P, Gefen D. Building effective online marketplaces with institution-based trust. Inform Syst Res 2004;15(1):37e59. [32] Gulati R, Gargiulo M. Where do interorganizational networks come from? Am J Sociol 1999;104:1439e93. [33] Datta P, Chatterjee S. The economics and psychology of consumer trust in intermediaries in electronic markets: the E-M trust framework. Eur J Inform Syst 2008;17(1):12e28. [34] Blau PM. Exchange and power in social life. New York, NY: Wiley; 1964. [35] Fukuyama F. Trust: the social values and the creation of prosperity. , New York: The Free Press; 1995. [36] Gefen D. Customer loyalty in e-Commerce. J AIS 2002;3(1):27e51. [37] Jarvenpaa S, Tractinsky N, Vitale M. Consumer trust in an Internet store. Inf Technol Manag 2000;1:45e71. [38] Luo X. Trust production and privacy concerns on the Internet: a framework based on relationship marketing and social exchange theory. Ind Mark Manag 2002;31(2):111e8.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002

K. Marett et al. / Technology in Society xxx (2014) 1e11 [39] Pavlou P. Consumer acceptance of electronic commerce: Integrating risk and trust with the technology acceptance model. Int J Electron Commer 2003;7(3):101e34. [40] Featherman M, Valacich J, Wells J. Is that authentic or artificial? Understanding consumer perceptions of risk in e-service encounters. Inform Syst J 2006;16(2):107e34. [41] Komiak SX, Benbasat I. Understanding customer trust in agent-mediated electronic commerce, web-mediated electronic commerce, and traditional commerce. Inform Technol Manag 2004;5(1e2):181e207. [42] Pavlou P. Institution-based trust in interorganizational exchange relationships: the role of online B2B marketplaces on trust formation. J Strat Inform Syst 2002;11(3e4):215e43. [43] Mallat N. Exploring consumer adoption of mobile payments e a qualitative study. J Strat Inform Syst 2007;16(4):413e32. [44] Charki MH, Josserand E. Online reverse auction and the dynamics of trust. J Manag Inform Syst 2008;24(4):175e97. [45] McKnight DH, Chervany N. Trust and distrust definitions: one bite at a time. In: Falcone R, Singh M, Tan Y-H, editors. Trust in cyber-societies. Berlin/Heidelberg: Springer; 2001. p. 27e54. [46] Deutsch M. The resolution of conflict: constructive and destructive processes. New Haven, CT: Yale University Press; 1973. [47] Barro R, Lee J-W. International data on educational attainment: updates and implications. Center for International Development, Harvard University; 2000.

11

[48] Chin WW. The partial least squares approach for structural equation modeling. In: Marcoulides GA, editor. Modern methods for business research. Mahwah, NJ: Lawrence Erlbaum Associates; 1998. p. 295e336. [49] Katz J, Aspden P. Theories, data, and potential impacts of mobile communications: a longitudinal analysis of U.S. national surveys. Technol Forecast Soc Change 1998;57(1e2):133e56. [50] Arnold M. On the phenomenology of technology: the “Janus-faces” of mobile phones. Inform Organ 2003;13(4):231e56. [51] Prasopoulou E, Pouloudi A, Panteli N. Enacting new temporal boundaries: the role of mobile phones. Eur J Inform Syst 2006;15(3): 277e84. [52] Prahalad CK. The fortune at the bottom of the pyramid: eradicating poverty through profit and enabling dignity and choice through markets. Upper Saddle River, NJ: Wharton School; 2005. [53] Ferneley E, Light B. Secondary user relations in emerging mobile computing environments. Eur J Inform Syst 2006;15(3):301e6. [54] Rossberg M, Schaefer G, Strufe T. Distributed automatic configuration of complex IPsec-infrastructures. J Netw Syst Manag 2010; 18(3):300e26. [55] Schoorman D, Mayer R, Davis J. An integrative model of organizational trust: past, present, and future. Acad Manag Rev 2007;32(2):344e54.

Please cite this article in press as: Marett K, et al., Using mobile devices in a high risk context: The role of risk and trust in an exploratory study in Afghanistan, Technology in Society (2014), http://dx.doi.org/10.1016/j.techsoc.2014.11.002