IT may lead to more rapid, better-informed decision-making by decreasing the cost of acquiring and generating information and reducing communication costs. Empirically, the effects of these cost reductions and better decisions should manifest themselves in enhanced firm productivity. Production function is widely used as the basis of estimating the measure of productivity. The production process of a firm would be represented by:
Qit = P(Kit, Cit, Lit, Rit, i, j, t)
It is assumed that this relationship can be approximated by a Cobb-Douglas production function.
Q = G(i, j, t) Kb1 Cb2 Lb3
Taking the natural log variants of production function are specified as follows:
lnQ = g(i, j, t) + b1 lnk + b2 lnc + b3 lnl + et
lnQit = bi + bj + b1 lnCit + b2 lnkit + b3 lnSit + b4 lnLit + et
where, i indexes the individual firm, t is the time, j represents the industrial sector, Q is output, C represents computer capital, K is non-computer capital, S is information services (IS) labor and L represents all other labor. The b’s are coefficients to be estimated, and e is the error term. Several variants of this basic equation are estimated. These alternative specifications relax the cross-equation equality restriction on the coefficients, allow for measurement error and the endogeneity of computer capital and are estimated on the basis of different sub-samples.
IT is a function of relative factor prices and output. To incorporate firm structure, it can be assumed that firm structure choices affect the overall level of IT, but do not change the responsiveness of demand to price. The transcendental logarithmic cost function, which is commonly applied in empirical production research, is given as:
Log Cost = Sai log pi SSaij log pi log pj + ao log O + S aio log pi log O
Where, K is ordinary capital with price pk, L is ordinary labor and O is firm output.
IT Cost Share = pcC/(pitC + pkK + plL)
Another common approach is to measure the correlation between computer spending and some performance metric, as suggested by equation:
Performance matrix = f(computers, environment, strategy, etc.)
Typical performance metrics used include (growth in) business profits, (growth in) sales, (growth in) market share, stock price appreciation, and various industry-specific measures. While there are numerous ways to operationalize such a relationship, they all seek to isolate the contribution of computers while controlling for other factors. If managers are rational, economic theory predicts that in equilibrium, high computer investors would not, on average, perform any better than low computer investors by these metrics. If significant correlations are found, they should be interpreted as indicating an unexpectedly high or low contribution of information technology, as compared to the performance that was anticipated when the investments were made. In addition, a key assumption of this approach is that businesses retain the value created by their investments in IT, or at least capture a significant portion of it.
The models are based on basic assumption that business value of information technology is either in the form of cost control or in terms of strategic gains.
This method uses the correlation between firm structure and IT capital to test the relationship between IT and coordination costs. Diversification and vertical integration are included as the dimensions of the firm structure. Diversification entails increased internal coordination costs because firms have to manage more complex and diverse activities. The impact of IT on vertical integration is determined by the degree to which IT changes the cost of internal coordination, external coordination, and production. The simplest empirical test is to calculate rank order correlations between three observable variables: IT capital (C), vertical integration (V), diversification (D). The equations are:
1a: V = mv0 + mvc C + other determinants of V + ev
1b: D = md0 + mdc C + other determinants of D + ed
1b: C = mc0 + mvc V + mdc C + other determinants of C + ec
where, ec, ed, ev are i.i.d error terms