An Investigation of Drivers and Barriers Stimulating in the Acceptance of Mobile Payment in Bangladesh

Bangladesh is one of the highest mobile phone users’ countries among ten in the world where mobile payment service is becoming a promising area with the speedy rise of mobile internet. Conversely, various researches on mobile phone users acceptance of mobile payment has not reveal clearly to a definite assumption for Bangladesh context considering technology acceptance model (TAM). A survey has conducted to collect data from 260 consumers of Bangladesh and then hypothesized model based on data analysis in SPSS. It has found that perceived ease of use and perceived usefulness are the drivers in the acceptance of mobile payment where subjective norms mediate them. Alternatively, perceived risk and perceived costs are treated as barriers toward behavioral intention to use mobile payment. This study explores new results in the TAM model in Bangladesh context and provided managerial insights to the mobile payment service providers to develop strategies in Bangladesh.


Introduction
Mobile payment (m-payment) is payment of train fare, parking, shopping and digital contents. Mobile payment services conducted via a mobile device -have been a key driver in socioeconomic development in emerging markets. Mobile payments is widely tipped to become a hot growth area for telecom business, and in some markets in Asia Pacific it already is , from smart-card-based mobile wallet services in Japan and Korea to remittance payment services in Southeast Asia. But it's early days, and the mobile payment pie is going to get a whole lot bigger. It has projected that global mobile payment revenue grow from $47.2 billion in 2011 to $998.5 billion in 2016 (Roberts, 2013) and forecasted value of global m-payment transactions to pass $958 billion by 2015 at a growth rate of 97% a year [74]. World is focusing towards developing countries to lead this mobile payment services. Kenya is leading the world in Mobile Money service through M-PESA since 2007 to allow microfinance-loan repayments [70]. It was then imitated by Philippine, India, Bangladesh and Pakistan.
In Bangladesh, bKash has succeeded to launch mobile payment service as part of mobile financial services in 2011 [19]. The service launched will facilitate faster transfer of money from almost every country to Bangladesh. Bangladesh is well headed in adopting new business of mobile phone and introduced Mobile payment with or without direct involvement of mobile network operators (MNOs). Mobile payment adoption appears strongest in Asia where Bangladesh and Pakistan have more than half of all mobile consumers registered their phones to make purchases [66].
Ten Banks and six MNOs have introduced Mobile payment service in Bangladesh and most of them emphasis on person to person (P2P) payment. Within one year after introducing Mobile Money, the market has 442,289 registered customers, 9,093 appointed agents and rolled TK. 20.71 million till March, 2012 [52]. Although people are more concern to P2P using Mobile payment service; there is a scope of more services permitted by Bangladesh Bank related to m-payment. When compared with the global leader Kenya, mobile financial services in Bangladesh has a lot of room to grow before it can claim any wide national impact. For instance, while 35% of Kenyan mobile money users use their accounts for advanced functions such as purchasing goods, receiving salaries and taking loans, in Bangladesh, mobile money use is heavily based on person-to-person transfers [35].
The growing rate of m-transfer and m-payment and the facilities provided by to Banks and MNOs will support enhancement of E-Marketing. As most of the people of Bangladesh are staying rural areas earning low amount of income and less educated mobile payment is not expanding with the availability of internet. People are now experienced to send/receive money as balance transfer, Bangladesh Post Office (BPO) service and mobile ticketing system. Now, it is the turn to develop merchant payment through using mobile payment service. But the trend of the transaction is very leisurely and it needs to minimize the gaps through positive mind set up towards mobile payment. For this reason the study needs to focus on finding drivers and barriers of mobile payment acceptance in Bangladesh. This research is highly attention to make a link those drivers and barriers with the impacts on mobile payment acceptance so that researchers and managers will have better outcomes to improve their research potentials or customers satisfactions.

Literature Review and Hypothesis Development
The mobile payment research has been streamed in two ways. It has measure as the mobile technology in constant evolution of mobile payment system [45]; others are concerned on the consumer acceptance of the technology with various empirical models based on Technology Acceptance Model (TAM), Theory of Planned Behaviour (TPB), and Innovation Diffusion Theory (IDT). Conversely, many empirical researchers bring out different antecedents of the mobile payment acceptance. This study reviewed several researches from 1999 to 2013 done in Europe, America, Asia, and Australia those have been developed a number of models attempted to understand factors influencing acceptance of mobile commerce. Among this TAM is mostly robust in explaining acceptance of mobile payment services. However, the original model has had to be revised to examine other aspects. This study explored that the same factors may be validated in different papers to have a positive, negative or insignificant effect on mobile payment acceptance. For example, ease of use is a strong determinant of behavioral intention [3] but it has been verified empirically to be insignificant in different studies [25] [85]. The effect of usefulness is mostly a significant factor of behavioral intention whereas it found insignificant at P<0.05 level [51]. What on earth are the factors that drive or obstruct the acceptance of consumers to use mobile payment in Bangladesh? This research attempts to examine this concern.
The number of studies on mobile payment acceptance in Bangladesh is very limited and most of the studies analyze concentrating on other service of mobile commerce. Furthermore, a few studies concern mobile payment overview in Asian perspective and most are concern on developed countries. There is no research found related to high concentration of m-payment and its acceptance in the context of Bangladesh. In this backdrop the main reason for undertaking the study is planned to identify and relate the drivers and barriers in the acceptance of mobile payment in Bangladesh.

Technology Acceptance Model (TAM)
The Technology Acceptance Model (TAM) is an information systems theory that models how users come to accept and use a technology [15]. The model had two main determinants which explain IT adoption: Perceived Usefulness and Perceived Ease of Use. Perceived usefulness is defined as being the degree to which a person believes that the use of a system will improve his performance. Perceived ease of use refers to the degree to which a person believes that the use of a system will be effortless. Several factorial analyses demonstrated that perceived usefulness and perceived ease of use can be considered as two different dimensions [23].
Davis reported that the relationship between perceived usefulness and adoption was significantly stronger than that of between perceived ease of use and adoption. Furthermore, he noted that perceived ease of use might even precede perceived usefulness, suggesting the existence of a causal relationship instead of the independence of the driving forces. Perceived usefulness is considered to have a positive effect on intention to use a technology. It is a measure of the subjective assessment of the utility offered by the technology.
Perceived usefulness is a significant determinant of attitude and intention [25] [14] and [71]. Perceived ease of use is the degree to which a person believes that adopting the system will be free of effort. Perceived ease of use has both an immediate effect and an indirect effect on adoption intention through mediating perceived usefulness [3] [34] [14]. Social and organizational variable such as subjective norm has shown direct impact on perceived usefulness and perceived ease of use. It is the most influential factor to mediate perceived usefulness and perceived ease of use to have direct impact on behavioral intention towards mobile payment use [63] [69] [67]. This research focuses on the key determinants of mobile payment acceptance and examined user acceptance considering behavioral intention and actual use in Bangladesh. Therefore, based on TAM, this study hypothesizes:

Perceived Risk and Cost
Perceived risk (PR) is universally acknowledged to be introduced by Bauer in 1960 [18] which refers to the extent to the prospective user expects the services to be risky. Perceived risk as a direct antecedent of consumers' intention to accept online transaction [57]. Perceived security risk is a vital determinant for mobile payment acceptance which has 6 contributing factors, namely, unauthorized use, transaction errors, lack of transaction record and documentation, vague transactions, concerns on device and network reliability and concerns on privacy [47]. It is also a critical factor in mobile commerce acceptance in Taiwan [39] [29]. There are 48% of the cellular users who refuse to use mobile payment services are due to security issue. Perceived risk is a key barrier to mobile payment acceptance in Bangladesh too.
Perceived cost refers to the extent to which people believe that use of mobile payment services will have monetary cost. It is considered as an important factor affecting user's behavioral intention to mobile payment [73] [47] and in similar domains like 3G mobile value-added services [37], 3G phones [81] mobile banking [56], and mobile commerce [29]. Consumers intend to accept mobile payment services when they perceived it as cost-effective. Consequently, the hypotheses on perceived risk and perceived cost will be: H 7 : Perceived risk has negative affect on behavioral intention of mobile payment in Bangladesh.
H 8 : Perceived cost has negative affect on behavioral intention of mobile payment in Bangladesh.

Research Model
The technology acceptance model shows that perceived usefulness and perceived ease of use has direct relation to behavioral intention and behavioral intention has direct impact on actual use. There are some other research found social influence or a subjective norm also has direct relation to both perceived usefulness and perceived ease of use.
This research tries to find out negative relation between barriers and behavioral intention too. So, perceived risk and perceived cost has negative impact on behavioral intention. Considering the above hypotheses, the research model of this paper is presented as figure 2. The perceived risks and perceived costs are added with the existing TAM theory to find out the impact in Bangladesh context.

Research Methodology
The study has covered selection of sample, nature & sources of data, developing data collections, interview schedules, pilot study use, and techniques and process of data analysis.

Questionnaire Development
The questions in the questionnaire were based on review of literatures and the specific characteristics of Bangladeshi people & market context. The specific characteristics represented different services and facilities provided by the MFS institutions and expected by the MFS consumers of Bangladesh. The questionnaire was translated into Bengali language to make clear and simple to understand by every respondent. The items in the questionnaire were constructed based on the consumer attitude and acceptance of m-commerce services through mobile payment.

Sample Size
The sample of the study were two mobile financial service (MFS) institutions & mobile network operator (MNO) and 260 Customers of Mobile Money service of m-payment using different MNO and MFS.

Data Sources
This study will considered survey of randomly selected users from different regions of Bangladesh. The respondents verbally replied to a structured questionnaire and their answers were recorded accordingly by the interviewers.

Preparation of Data
The research used one set of questionnaires (structured) for the customers using Mobile Payment services in Bangladesh. Five points Likert Scale used in preparing the study. The questionnaires set up in accordance with External Quality Assessment (EQA) criteria in order to be able to compute scores. For each criterion several questions need to be selected to obtain a valid representation of each individual criterion. Pilot Study has done for finalizing the survey schedule through discussion with the expert users of mobile payment services.

Data Analysis
As far as scale-based variables were concerned, principal factor analysis was performed. In this study, the correlation matrix was used to obtain Eigen values. To facilitate the interpretation of factor loading, VARIMAX rotation was performed. In addition, Pearson correlation analyses were also conducted to examine the relationships among the constructs. SPSS software version 20 was used for analysing the data.

An Investigation of Drivers and Barriers Stimulating in the Acceptance of Mobile Payment in Bangladesh
The study were investigated demographic characteristics in regards to the respondents' gender, age, occupation, income, residence region, e-commerce service use, mobile network operator use, mobile financial service institute use, and number of monthly transaction. Missing values, outliers, and distribution of all the measured items were conducted to avoid the error of the estimates and also to purify the data.
The male and female ratio of the respondents was 69.2% and 30.8% respectively. It indicates that male respondents are using online shopping services more than female. Most of the respondents were at 18 to 25 years age range and 74.2% are students, which imply that mostly youth level students are using e-commerce services. Monthly income of the respondents is mostly less than BDT 10000 (56.5%) and their residence regions are mostly Khulna (28.8%), Rangpur (20.4%), and Rajshahi (16.6%). The e-commerce services using mobile money are mostly mobile top-up (39.6%), send money (35.4%), and utility service (10.8%). Respondents are mostly using mobile financial service institute account of either DBBL (48.1%) or bKash (44.2%) to use mobile money for avail e-commerce services. Those who are availing e-commerce through mobile money are making five or less transactions (86.9%) in a month.

Reliability and Validity
There are twenty four items have been considered for the study. Multi-item scales (Six Point Likert Scale) response format has been used to operationalize each individual item of the questionnaire. The overall measurement for contextually formatted items is described in Table 2 result shows that scales used for the data collection are mostly reliable with highly recommended alpha scores which is within the range of good scale reliability [54].
The study identifies 33 items based on Eigenvalue -1 or above includes into the list of items. Statistically 24 items construct seven factors which can explain 64.245% of the field. So, this study considers these factors are important for the study. Factor-1 explains 23.281% having Eigenvalue of 5.587and Factor-7 explains 4.265% with Eigenvalue 1.024 as lowest (Table 4). Hence, factor-1 is the most important in the current study which is related to behavioral intention to use mobile payment in Bangladesh.

Factor Analysis
The results obtained from 260 respondents had been thoroughly analyzed. Applying SPSS, the Principal Component Analysis (PCA) was carried out to explore the underlying factors associated with 33 items. The constructs validity was tested applying Bartlett's Test of Sphericity and the Kaiser-Mayer-Olkin (KMO) measure of sampling adequacy analyzing the strength of association among variables. Result of the Bartlett's Test of Sphericity and KMO revealed that both were highly significant (0.823) and eventually concluded that these variables was suitable for the factor analysis (Table 3). Deciding upon the number of factors that can be retained is difficult but initial runs-based on Eigenvalues showed 7 factors explained 64.245% of total variance (Table 3).  To determine the minimum loading necessary to include an item in its respective constructs, loading 0.50 or greater are very significant [21]. For this study, the general criteria were accepted items with loading of 0.506 or greater.
The factor analysis revealed seven factors affiliated from 24 items (Table 5). Generally, the factor is the natural affinity of an item for a group. The higher factor loading indicates the stronger affiliation of an item to a specific factor. The findings of this study indicate that each of the seven factors: behavioral intention (BI), perceived ease of use (EOU), actual use (AU), subjective norms (SN), perceived usefulness (PU), perceived risk (PR) and perceived costs (PC) were homogeneously loaded to the different factors. Each of the items that loaded into nine different factors has proven as significant related to the mobile payment users' expectation.

Multiple Regression Analysis
The performance of regression analysis should be justified with the normality of distribution and no presence of multi-collinearity among the indigenous variables. The normality has been checked through histogram while the multi-collinearity has been examined through tolerance level and VIF value of regression statistics. The tolerance level and VIF value justify the absence of multi-collinearity among the predictors [5] [30]. In this study VIF value appeared between 1.000 to 1.155 in four regression models. Again, the tolerance level appeared from 0.844 to 1.000 in five regression models ( Table 6). These indicators prove that the predictors in the multiple regression analysis are free from the multi-collinearity. Thus the regression model run should be directly administered for examining the degree and magnitude of the effects of variables. Again, in model 2 four characteristics (perceived ease of use, perceived risks, perceived costs and perceived usefulness) considered in the model account for 37.6% mobile payment use intention by the consumers in Bangladesh. In model 3 two characteristics (perceived ease of use and subjective norms) considered in the model account for 32.3% on perceived usefulness; and subjective norms in model 4 considered 34% on perceived ease of use by the mobile payment consumers in Bangladesh. The study showed that eight hypotheses were supported out of eleven hypotheses significant at p < 0.05 (Table 7).

Conclusions and Implications
This study was attempted to examine and understand the consumers' behavioral intention and use behavior towards mobile payment services in Bangladesh. The mobile payment service providers are characterized by the involvement in making awareness and educating to use their mobile payment services. Presently, they are highly competing for providing diverse mobile payment services along with network quality and network extension for making network loyal customers. According to this study, behavioral intention is the highest significant factor to influence actual use of mobile payment for the consumers in Bangladesh. Again, perceived risk is the highest significant factor to influence behavioral intention following perceived costs and perceived ease of use. Finally, subjective norms affect significantly to mediate perceived usefulness and perceived ease of use to influence behavioral intention to use of mobile payment services in Bangladesh.
The primary theoretical contribution of this study is the identification of factors that can be used to enlighten and foretell consumers' intention and use of mobile payment services, particularly within the Bangladesh context. This research also creates a research interest on mobile payment for the business researchers. Again, perceived risk and perceived costs have been found as strong factors of TAM model relevant to mobile payment services in Bangladesh. Finally, it has found that people are very much interested to involve in mobile payment use those are day to day business to customer (B2C) payment like transportation ticket purchase, utility bill payment, money transfer for personal & business purposes, and online shopping of home appliances. So, the findings of the study can help mobile payment services providers in their operation and strategic plan of marketing of mobile VAS applications.

Limitation and Future Research
The sample size of this study was not necessarily representative of the Bangladeshi population as a whole as it ignored same number of population from different regions. Secondly, the generalizability of this research may be impacted by fact that the sample's is skewed towards males. In terms of future research, a large scale study with more representative sample could be conducted to measure the degree of impact of the different drivers and barriers in the acceptance of mobile payment to enhance the generalizability of the research conclusions.
Mobile payment service is very new to Bangladeshi consumers and it is still at early stages in Bangladesh. The current target market for mobile payment customer is relatively small due to the lack of expanded use of mobile payment in online merchants, banks, MFS, and MNOs. There is a good potential for introducing mobile payment service in Bangladesh. This research has served to enhance the understanding of the impact drivers and barriers to influence consumers' acceptance towards mobile payment use it in the context of Bangladesh. It has demonstrated that there were multiple impact factors at work for influencing the consumers and that some are more influential under given circumstances. The knowledge gained by this study into the motivators and inhibitors of mobile payment is useful for practitioners who aim to maximize consumers' interest and business expansion at mobile payment use of B2C services in Bangladesh.