Night-Shift Turnover at Cedarline Components
Cedarline Components runs two injection-molding shifts in a 120-employee plant. Night-shift operators turn over at 48% a year versus 12% on days; there are 54 night operators. Starting pay is $17.50 an hour against a local market of $18.75, and schedule changes are posted as little as 36 hours ahead. Exit interviews cite unpredictable schedules and no visible path to higher pay. Last quarter the plant paid $86,000 in overtime and $31,000 to a temp agency, and scrap on nights runs 4.1% versus 2.3% on days. Training a new operator costs about $4,200, and recruiting and onboarding add roughly $3,800. The plant manager proposes raising night pay to $18.75, a $140,400 annual increase. The controller argues that a fixed two-week schedule plus a skill-based pay ladder costing $95,000 a year would be cheaper. The owner wants a recommendation before the next budget cycle.
1. Estimate the annual cost of night-shift turnover and compare it with the proposed pay increase. Does the raise pay for itself?
Read the worked answer
Separations: 54 × 0.48 ≈ 26 per year. Replacement cost per operator: $4,200 training + $3,800 recruiting/onboarding = $8,000. Annual turnover cost ≈ 26 × $8,000 = $208,000, before overtime, temp labor, and scrap.
Proposed raise: 54 × 2,080 hours × $1.25 = $140,400. If turnover fell to the day-shift rate (12%, about 6.5 separations), savings would be about 19.5 × $8,000 = $156,000-roughly enough to cover the raise, but only under a best-case assumption that pay alone changes behavior. Exit interviews point to schedule instability and lack of advancement, which a wage increase does not fix. The $86,000 overtime and $31,000 temp spend may partly reflect vacancies, but not all of it is avoidable turnover cost, so the controller is right to be skeptical of a pay-only case.
2. Recommend an intervention package, justify it with motivation theory, and explain how to phase it so the owner can evaluate results.
Read the worked answer
Use Herzberg's two-factor logic: $17.50 against an $18.75 market rate is a hygiene/equity problem that creates dissatisfaction, but schedule predictability and a visible skill ladder are the motivators that affect whether people stay. Adams's equity theory also predicts that workers who see themselves underpaid relative to market will reduce effort or leave, so some pay correction is needed. A bundle is stronger than either proposal alone: (1) post schedules 14 days ahead and limit last-minute changes, with a premium for voluntary changes; (2) a skill-based ladder with defined competencies and raises at each level; (3) a smaller night differential, such as $0.50–$0.75; (4) cross-training and a shift-swap system that gives operators more control.
Phase it on nights for six months rather than plant-wide. Fund the schedule system and ladder first (about $95,000), add the partial differential (about $56,000–$84,000), and hold the full $1.25 raise in reserve. That keeps first-year cost near the pay-only option while attacking the stated causes.
3. What data should be collected during the pilot, and what decision rule should trigger expansion?
Read the worked answer
Collect monthly: separations and 90-day new-hire retention; number of schedule changes per operator and days of notice; overtime hours and temp-agency spend; scrap rate by shift; absenteeism; internal promotions and ladder progress; and a short pulse survey on schedule fairness and perceived advancement. Include cost per good unit so a scrap improvement is not hidden.
Decision rule: expand only if, after six months, night turnover falls below 25%, 90-day retention rises above 75%, schedule changes fall by half, and scrap moves toward 3% or better-while total labor cost per unit stays at or below the current baseline. If turnover improves but scrap and overtime do not, the problem is supervision or equipment, not scheduling. If nothing moves, stop the ladder expansion and revisit the wage gap with fresh market data.