I in the manufacturing sector was examined by Loveman (1994). The results indicated that "the marginal dollar would best have been spent on non-IT inputs into production, such as non-IT capital" (Loveman, 1994, p. 27). Another study compared the productivity of information and production workers (Roach, 1991). One finding of this study was that information worker productivity either declined in some sectors or failed to keep pace with production worker productivity in the manufacturing sector.
A study by Weill (1992) classified EDI into three categories based on the management goals supported by the system. This study identified significant productivity gains that result from transactional EDI, but did not find any positive impacts for strategic or informational (infrastructure) EDI systems. A study by Barua, Kriebel, & Mukhopadhyay (1995) proposed a two-stage analysis of intermediate and final output variables for measuring EDI contributions in the manufacturing sector. This study found that EDI had a significant impact on intermediate variables such as capacity utilization, inventory turnover, and product quality, but found little impact attributable to EDI on return on assets or market share.
The contradictory findings on the value of EDI with respect to organizational performance are "symptomatic of the difficulties of conducting such an analysis" (Mukhopadhyay, Kekre, & Kalathur, 1995, p. 139). The majority of studies examining the value of EDI to organizational performance have not found strong evidence of EDI impactùeither positive or negativeùon organizational performance.
A study at LTV Steel on its EDI linkages with outside processors established that, compared to the company's prior manual mode of information exchange, EDI transactions had more favorable effects on quality and inventory measures (Kekre & Mukhopadhyay, 1992). A study of an automobile manufacturer and its first-tier suppliers found that the...