Behind the Diffusion Curve : An Analysis of ATM Adoption

The paper examines the impact of firm characteristics, market structure and state regulations on the adoption of ATMs by banking organizaitions. A grouped duration data framework is used to investigate the effect of these factors on the hazard rate of adoption. The analysis shows that larfer fimes, especially those owne by bank holding companies. Were earlier adopters. At an given times, the propertopm of prior adopters in a market increased the likelihood that a non-adopter would do so. Market concentration increased the hazard rate if only a small proportion of firms in the market were using ATMs. Firms operating in utban, high wage markets and experiencing relatively rapid growth of deposits, especially demand deposits, were more likely to introduce ATMs. The mandatory sharing of ATM systems required by some states had a negative, though not significant impact. However, the conditional probability of adoption was higher in states where branching was either prohibited or restricted and off-premise ATMs were allowed. The study also shows that the effect of the explanatory variables changed across different phases of ATM diffusion. Further, it is demonstrated that not accounting for the grouping of the data and/or the sue of restrictive parametric hazards, as in some past analyses, affects both the estimated coefficients and their standard errors, and may even lead to incorrect qualitative conclusions regarding the impact of certain covariates