Output vs Input Inventories. The literature on inventory investment can be generally classified into two categories. One focuses on the use of final-good inventories for the purpose of stock-out-avoidance, and it has received abundant attention. The other, which had mostly been neglected, incorporates the stage-of-fabrication linkages within and between firms where input inventories are uti- lized for production-smoothing. Input inventories, conventionally defined as raw materials and work-in-process products, arise whenever there is a gap between the delivery and use of inputs. They had been neglected in earlier models because their importance was largely observable in durable goods, and earlier models commonly excluded such goods7. However, the significance of input inventories in production and business dynamics has been pointed out in recent work. Xxxxxxxxx et al. (2001), for instance, show that input inventories are larger and fluctuate more than finished-good inventories in US manufacturing. Findings in Xxxxxxx and Xxxxxxxxx (2005) show that at the disaggregated level, input and output inventories contributed differently to the reduction in output volatility during the Great Moderation. Xxxxxxxxx (2005) states that since the usage of input materials is a factor of production, decisions on production-smoothing and output-inventory are inherently related to input-inventory decisions. Eroglu and Hofer (2011) investigate the contribution of the different inventory types to a firm’s financial performance, and find that Raw Material inventories contribute the most compared to Work-in-Process and Final Good invento- xxxx. Meanwhile, Xxxxxxxxx and Xxxxxxxxx (1999) uncover a negative correlation between productivity growth and input inventories, particularly Work-in-Process (WIP). Their key argument is that reducing WIP exposes production problems on the shop floor, enabling 7See Blinder and Xxxxxxx (1991) for a summary of earlier literature. them to be attended to hence boosting productivity. In turn, an increase in productivity leads to a further reduction in the need for WIP. From these publications arises the motivation to disintegrate my inventory variable into its different components, and investigate whether they have different effects on the relationship between hours and productivity. The US manufacturing data on input inventories is only available for 2-digit level industries. I run the FASVAR second step, with total inventories replaced, in turns, by Final Goods (FG), Work-i...
