Uneducated guesses about the current state of algorithmic labor management systems

Things I suspect are true, but don’t know yet:

  • A large percentage of workers in the USA doing work that is planned by the hour have their schedules determined by a small number of large, complicated, semi-automated systems.
  • LaborPro is one of those systems.
  • There are others, but there aren’t that many others, because the barrier to entry for creating scheduling software is high. Scheduling is a very persnickety part of computer science, and it’s not very intuitive.
  • When you add on that the main concern of these software systems is compliance with a pastiche of state and local labor rules, the problem domain goes from “persnickety” to “fiendishly complicated.” The systems must be constantly updated, and the updates must be done by people that are skilled in this area.
  • These systems are primarily — almost exclusively — concerned with two things:
    • Minimizing the overall cost of labor, and
    • Minimizing the risk of expensive fines for being out of compliance with labor regulations.
  • These systems might be good at creating schedules that are in compliance with regulations, and these systems might be good at reducing overall cost of labor, but these systems are NOT good at coming up with schedules that take workers’ needs and desires into account.
  • In fact, though many parts of these systems are devilishly complex, the parts having to do with making “good” schedules are childishly naive.
  • But it’s hard to uncover that, because the overall system is so complex, and these systems are very competitive; they are not inclined to share or collaborate on their scheduling systems.
  • In addition, it has always been a fundamental part of the “labor vs management” divide that management has lots of information, and labor has very little information. “Don’t discuss your wages!”
  • So not only are these systems not very good at coming up with “good” schedules, they are very disinclined to come up with a clear definition of what “good” looks like, and they won’t share what (if anything) they are doing to balance the cheapest schedule with the best schedule from a worker’s perspective.
  • A good first step is to define what a “good” schedule is from a worker’s perspective. Five years spent working in a UX consultancy has taught me that listening to the end-user is done SURPRISINGLY RARELY. These systems are probably exquisitely attuned to the needs of the people paying for the construction of the software — but can they articulate how they are working to create “better” schedules for workers? I don’t know, and I’d like to find out!

Possible next steps are to go splash around in the work that the Oxford Internet Institute is doing with gig economy work. Also to look up LaborPro training and see if I can find out the tools a shift manager uses when putting together their shifts for the coming weeks and months. What does the system suggest? What tools are made available to the manager to amend, update, and override the suggestions?

And most of all, what does GOOD look like? Schedules that are predictable — the same or similar from week to week? Schedules that are contiguous — offering a smaller number of bigger chunks of hours (when would this be a good thing, and when would this be a bad thing?) Schedules that provide the number of hours so that the worker can hit their benefits thresholds?

This is all really disorganized at this point, especially because some of the work that is being done is around the ethics of the gig economy, and I’m most interested right now about the ethics of scheduled part-time workers in “normal” jobs.

Do you happen to know something about this domain? Who are the authors I could or should be reading? Who are the lawyers that specialize in not only workers’ rights, but advocate for workers’ quality of life?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: