A study of the impact of information technology on firm performance: a flexible production function approach

Most previous studies of the impact of information technology (IT) on firm performance have relied on the Cobb-Douglas production function, which makes unrealistic assumptions. These include assuming that (1) the scale elasticity, i.e. returns to scale, is constant throughout the entire output range; (2) the elasticity of substitution between inputs is one, whereas it can be anywhere from zero to infinity in practice; and (3) there are no complex interrelationships between inputs, i.e. no squared or product terms. The translog production function (used in two previous studies) does not make these assumptions and is therefore more accurate. However, the two previous studies found no significant difference between the two. This research has extended previous work on the impact of IT investments by providing more accurate results in some areas and new findings in others. IT's importance is shown to be much greater than previously thought. The translog function is used to demonstrate that the productivity paradox does not hold (i.e. both IT labor and capital indeed have a significant impact on firm revenues), but with considerably different results, i.e. the impact of IT is markedly more significant than that obtained from the Cobb-Douglas. Another interesting finding is that of scale elasticity. Our results show that this is less than 1, implying decreasing returns to scale of output. Finally, the substitutability between inputs was determined, which is not possible with the Cobb-Douglas. It was found that the partial elasticities of substitution for IT labor with IT capital and non-IT labor were quite substantial.