Using Heckman’s sample selection methodology to analyse individual use of Internet banking we find a negatively significant Inverse Mills ratio. This is consistent with other studies. We go on to argue that this indicates that the probability of using Internet banking declines with the marginality of the Internet user. We further link this marginality to low cost Internet access and thence trust in this access. The inverse Mills ratio squared is also significant, thus both questioning the assumption of a bivariate normal distribution between the error terms from the selection and Internet banking equations, and also suggesting that there is a nonlinear relationship between Internet user marginality and the probability of using it for Internet banking.
|Name||Bath Economics Research Working Papers|
- Marginal user
- sample selection
- Internet banking