A partial adjustment approach to evaluating and measuring the business value of information technology

In this paper we develop a partial adjustment approach, based on the theory of partial adjustment, to assess the value of information technology. We also propose a performance measure derived from the approach. The constrained (structural or implicit) and unconstrained (non-structural or explicit) models derived from the approach are fitted into a set of panel data at the country level, covering the period from 1993 to 2006 for a sample of 12 countries including the G7 nations; and are estimated by the seemingly unrelated regression estimation. The results produced by the models are robust with respect to the choice of the production function representing the true (desired or maximum) output and challenge the widely held belief that the productivity paradox has disappeared and the view that the productivity paradox exists in developing economies only. The proposed approach compares favorably with the time-varying stochastic production frontier approach since it is easier and simpler to apply. But, unlike the non-parametric data envelopment analysis approach, it is parametric.