Fire At Random
Note that since teacher compensation costs increase as a function of experience, LIFO is actually worse than the equally objective practice of firing teachers at random. LIFO maximizes layoffs relative to financial targets. Doing layoffs by lottery would allow districts to fire fewer teachers.But of course doing layoffs by lottery would be a pretty silly way to run an organization.[emphasis Sara's]
The crazy thing here is that, while Yglesias is offering layoffs by lottery as a ridiculous example, the reality is that doing layoffs by lottery is actually a totally real practice defined in teachers union contracts.
I recently spent more time that I would have liked reviewing teacher union contracts from several school districts in a Pacific Northwest state. Each of the contracts I reviewed spelled out a basically common practice for teacher layoffs:
- Select the "employment categories" in which teachers are to be laid off. This basically means that you decide how many elementary classroom teachers, elementary special teachers, secondary teachers in each particular subject, etc. you need to layoff.
- Rank-order teachers based on seniority. The least senior teachers get laid-off first. Most contracts also stipulate that all "provisional" or "non-permanent" [read: non-tenured] teachers in the category get laid off before anyone else does (or sometimes this is not in contract but in state law).
- In the event of a tie based on seniority, the teacher(s) who have lower amounts of advanced higher education credits will be laid off first. [Note, this is an especially stupid criteria because, while there is some evidence showing returns to teachers' years experience in the first several years teaching—so the group of teachers most likely to be on the layoff chopping block—there is no evidence that accumulating more advanced postgraduate credentials improve teachers' effectiveness, except in the case of secondary math and science teachers earning advanced degrees in those specific subjects. Note that this criteria also has the same perverse effect as the LIFO criteria, in that because teachers with more coursework get paid more, it results in more teachers needing to get fired to reach the same budget target.]
- In the event of a tie in both seniority and higher education credits, teachers to be laid off will be selected by lot. In other words: by lottery. In other words: totally at random.
Note that performance is included nowhere in here. Apparently, considering teacher performance—by any measure—in layoff decisions is so dangerous that it is better to make layoff decisions completely at random than to even attempt to incorporate these deeply flawed and dangerous performance measures into layoff decisions.
One might scoff that oh, a tie in both seniority and higher education credits must be rare&jmdash;these random layoffs can't really happen. I admit that I don't have any solid information on how often this occurs (given that teacher compensation, tuition reimbursement, and state licensure policies often combine to incentivize teacher coursework in ways that make it likely teachers with similar experience will have similar numbers of credits, and that there are some BIG districts out there, I'd guess more often than you'd think). But, I have heard enough anecdotal reports to be convinced it happens at least sometimes. Which seems like more than it should ever happen.
Even if you believe that any judgment we could make about teacher performance is inherently flawed/unreliable/subjective, can't we at least agree that even a really, really crappy measure would be slightly better than making decisions totally at random?