When I worked in consulting we had a large retail client who had a lot of turn-over and was looking for ways to "stabilize" hiring. (Short of paying more money, of course!) When we looked into it their system was regularly over-scheduling shifts "just in case" there was a rush and then having store managers "cut" shifts if there wasn't a rush. There aren't a lot of surprise "rush" shifts in shopping malls outside of the holiday season so sometimes employees would come in, work an hour of a scheduled eight hour shift, and then get cut. The cuts tended to happen by "seniority" (so newest hires, paid the least) and a decent number of these employees who were frequently cut were traveling at least an hour by public transportation. Some days employees lost money by going to work.

We were thrilled by what we'd found! It would be a quick and easy fix for the client: they had solid and reliable sales forecasts for each location so estimating the number of employees they'd need on a given day wasn't going to be difficult. In fact, they were more or less already doing it. It would hardly cost them anything to implement aside from a few tweaks to the corporate scheduling tool and a bit of training for the store managers. So when management immediately shot down our proposed solution it came as a bit of shock. The erratic scheduling and the "cut" shifts were a feature, not a bug, we were told. It kept employees from getting a second job that they might dare prioritize over their current job.