just-in-time inventory effect: Does it hold under different contextual, environmental, and organizational conditions?, The

Journal of Business Logistics, 1998 by Droge, Cornelia, Germain, Richard

Just-in-time (JIT) refers to a collection of practices that eliminate waste. These organizationwide practices encompass the entire logistics flow of materials from purchasing through production and distribution. The elements of JIT may include shared product design with suppliers and customers, movement toward single sourcing, proximate suppliers and customers, reduced machine set-up times, total preventive maintenance, reliance on analytic tools (such as fishbone diagrams) to identify sources of defects in products and processes, demand-pull support, and cellular plant layout, among others.'

The benefits are pervasive and include lower total system costs and improved product quality. A good deal of anecdotal evidence is available to support the claim of improved performance derived from adoption of JIT components. Some plants have reduced in-process inventory more than 50% and lead time more than 80%. The purpose of this research is to examine empirically the relationship of JIT adoption with one dimension of performance: the level of inventory held by the firm. References to JIT's ability to reduce inventory buffers or slack resources (that is, inventory levels) abound in the literature and support our claim of widespread belief in an inventory effect from JIT.2 The initial research question addressed here is whether the level of JIT correlates inversely with the level of inventory.

This straightforward question, however, fails to capture the complex workings of modern corporations. For example, organizations that are small relative to customers may not realize an inventory effect from JIT adoption because customers push inventory back onto them. To address these and related issues, we also ask whether the level of JIT correlates inversely with the level of inventory after controlling for the linear (or nonlinear) effects of context, environment, and organization. Figure 1 illustrates the framework of the study. The upper part depicts the initial thesis of an inverse relationship between JIT and inventory. The lower part presents the context, environment, and organization control variables. Context includes firm size, size relative to customers and suppliers, the number of SKUs, and production technology (or production routineness). Job shop production (that is, short runs, many products) and continuous process production (long runs, few products) exist at opposite ends of the production routineness continuum. Environment includes the rate of change in demand, products, and processes, as well as industry growth and concentration. Organization includes organicity and executive tenure. An organic organization is more decentralized, specialized, integrated, and technocratic than its counterpart, the mechanical firm. In a typical organic firm, teamwork replaces functional silos, information more readily passes throughout the firm, lateral dominates vertical communication, decision making is pushed down the organizational chart closer to relevant information possessors, and specialized skill levels increase. As implied by the metaphor, organic firms are more flexible than mechanical firms, a more appealing property when the environment is dynamic. First, examine whether the inverse relationship of JIT with inventory holds when controlling for these variables linearly (that is, using partial correlations). Second, nonlinear effects are studied using subgroup analysis; for example, does the inventory effect hold for small versus large firms, for firms that are smaller than customers versus larger than customers, for firms that are organic versus mechanical, and so on? The analysis of the inventory effect of JIT while controlling for the linear and then nonlinear effects of context, environment, and organization will enrich logistics managers' understanding of the inventory effect of JIT by illustrating under which conditions JIT improves the inventory performance of the firm.

In the section that follows, we delve more deeply into why JIT should associate with less inventory and broadly explain why it is important to control for the linear and nonlinear effects of context, environment, and organization. We discuss a survey of 200 U.S. manufacturers and then present and discuss empirical results.

BACKGROUND

The Inventory Effect of JIT

JIT's inventory effect results from three interrelated processes. The first concerns a focus on product and process quality. The adoption of analytic tools (such as fishbone diagrams, Pareto charts, and design of experiments) and the inclusion of direct labor into quality improvement programs have a significant effect on inventory levels. Analytic tools are used to spot the sources of defects, whether these sources originate in defective assembly processes, materials, or materials design. These tools become all the more powerful when their use is orchestrated by direct workers in production and logistics whose expertise and knowledge is tapped by quality circles. As sources of defects are identified and remedied, product quality improves, the number of quality control inspectors declines, and, more important, the amount of scrap and rework inventory declines. Reductions in work-in-process inventories lead to a lesser need for safety stock based upon expected defect levels in inventory. Safety stocks of finished goods may also decline because the expected defect rate is lower. Since products are of a higher quality, the amount of after-sales service repair work declines, leading potentially to reductions in the amount of spare parts kept on hand.3