Perceived risk and Internet banking

Bankers and consumers are both interested in the potential for Internet banking. Individuals have been adopting die Internet in large numbers, with more than half of all American households having some form of Internet access by 2000. Banks too have been developing their infrastructure to address what they perceive as a growing demand for online services, with 84% of all accounts offering some form of Internet banking by 1999. However, the adoption rate has not followed the hype. By 2000, the proportion of households using Internet banking was less than 10%. This research looks at the critical factors needed to promote banking adoption from the consumer's perspective. We use a consumer utility maximization framework, and include in the consumption bundle the possibility of using conventional, phone-banking and/or Internet banking. Phone-banking is added because it could be seen as a substitute for Internet banking. Many of the same services are available on both, and many of die restrictions are the same, i.e., no cash can be withdrawn from either.

Using the utility maximization approach, we are able to conclude that adoption depends on marginal utility gain, marginal cost and a risk premium; where risk premium is the product of subjective probability of adverse outcomes from the technology and the utility of each adverse outcome. We use logistic regression to explore what factors are important to consumers adopting Internet banking in general. A conditional logit model is used to estimate the sensitivity of different decision factors to the marginal propensity of phone-bank customers adopting Internet banking and vice-versa.

The overall utility maximization model is consistent with our results from these logistic regressions. The results presented in this research also support the hypothesis that the subjective probability of security problems experienced by Internet bank customers is not the same for all customers, and that it depends on their level of education. Varying subjective probability means that the risk premium can be affected by exogenous factors, in this case education. In other words, banks could affect the risk premium of their customers, thereby affecting adoption rates.