Product reliability is a key concern for manufacturers. We examine a significant but under-recognized determinant of product reliability: the rate of workers quitting from the product's assembly line, or its worker turnover. While modern manufacturers make extensive efforts to control defects and assure quality worksmanship, some quality variation in the manufactured units may be revealed only after they have been used repeatedly. If this is the case, then the disruptiveness of high turnover may directly lead to product reliability issues. To evaluate this possibility, our study collects four post-production years of field failure data covering nearly fifty million sold units of a premium mobile consumer electronics product. Each device is traced back to the assembly line and week in which it was produced, which allows us to link product reliability to production conditions including assembly lines' worker turnover, workloads, firm learning, and the quality of components. Significant effects manifest in two main ways: (1) In the high-turnover weeks immediately following paydays, eventual field failures are surprisingly 10.2% more common than for devices produced in the lowest-turnover weeks immediately before paydays. Using post-payday as an instrumental variable, we estimate that field failure incidence grows by 0.74-0.79% per 1 percentage increase in weekly turnover. (2) Even in other weeks, assembly lines experiencing higher turnover produce an estimated 2-3% more field failures. We demonstrate that staffing and retaining a stable factory workforce critically underlies product reliability and show the value of connected field data in informing manufacturing operations.
Keywords: Data-driven workforce planning, Empirical operations management, Employee turnover, People