Let’s break the Richmond numbers out by school.
You read that right: 94% of the teachers at Lucille Brown were absent from work more than ten days in 2014!
For sure, the wide range here tells us we have (well, had in 2014) serious management problems in many of our schools: We’re paying a lot of teachers for not working. Of course, our School Board does not tell us about such problems or whether it is demanding that the the Superintendent do something about them.
Looking at the schools where the rate was >65%, we find half of the eight high schools (Armstrong, Jefferson, Open, and Wythe) (counting Franklin as both a high and middle school), two thirds of the nine middle schools (Binford, Brown, Elkhardt, Hill, King, and Thompson), and 15% of twenty-six elementary schools (Fox, Chimborazo, Fairfield, Mason).
The preponderance of our (awful) middle schools at the not-at-work end of the list raises the obvious question: Does the absence of all those teachers affect performance. After pulling the 2014 SOL scores and juxtaposing them with the absence rates, it appears that the short answer is “no.”
For the longer answer, let’s start with the elementary schools.
This plot of reading and math pass rates v. the teacher absence rates shows essentially no correlation between the two variables.
Aside: Correlation is a necessary but not sufficient requirement for causation. If the scores and absence rates correlated, it would suggest but not prove that the teachers’ absence was affecting the pass rates; the absence of a correlation, however, tells us that the variables are not causally related.
Well, on to those middle schools.
(Note, I’ve again included Franklin in both the middle and high school categories because it has students at both levels. It would make about the same amount of sense to leave Franklin out for failure to fit either category.)
As you see, we have higher absence rates and lower pass rates but no statistical relationship.
Last, the high schools.
Here, at last, some slight correlations (ca. 10% on R2, i.e., R of about 0.3).
If we delete Franklin, the correlations improve a bit, but still not to anything dramatic.
Note that reading goes the wrong way: If the correlation were better, we could wonder whether reading performance might be improved by the absence of the regular teachers. But we’re stuck with the numbers at hand.
That leaves the question of financial consequences, the cost of all those substitute teachers.
If Richmond could cut substitute teacher use by half, it could free up a couple of megabucks to give a 2% raise to its teachers (for actually working).
I’ll bet you a #2 lead pencil it doesn’t happen in your or my lifetime.