This paper uses a unique data base of plant-level cross-sectional data to analyse the relative performance of Just-in-time (JIT) plants operating in two distinct manufacturing industries: electronic-components and auto-parts. The multivariate tests show that JIT plants use significantly less work-in-process and finished goods inventories than do non-JIT plants. JIT plants are significantly more profitable than non-JIT plants. Furthermore, JIT plants have significantly smaller total and variable costs than do non-JIT plants. Fixed cost differences, on the other hand, are not significant. JIT plants exhibit significant learning curve effects in that the earlier adopters use less inventories and are more profitable than later adopters. JIT plants that claimed to be more successful at controlling process quality also tended to have significantly less work-in-process inventory, less total costs and more profits. On the other hand, JIT plants that claimed to be more successful at controlling product quality were not significantly different from other plants. Finally, unionization appears to have no impact on the relative performance of JIT versus non-JIT plants.
