The Cranky Taxpayer 
Facilities I 

In October, 2002, the BCWH
architecture firm submitted its Facility Master Plan for the Richmond Public
Schools. RPS formerly had the executive summary on its
Web page. They have since
removed it. Sigh. The Executive Summary summarizes the recommendations of the report. To get the raw data, however, you have to struggle through the three volumes of the Plan itself. There are data there that help understand where all the money is going. In short, we are wasting about more than $15 million a year in the instruction budget, mostly in the elementary and middle schools and we are paying for about 4% excess capacity in elementary schools and 19% (!) excess in middle and high schools. The first numbers that fall out of the Plan are the enrollments vs. the capacities of the schools. Our schools are heavily underused. Among the mainstream high schools, Marshall, Armstrong, Kennedy, and Jefferson all are using less than 80% of their capacity. In contrast, Huguenot is nearly 20% over capacity. Among the middle schools, Henderson and Mosby are using only about 60%: All told, we have 2859 unused places for middle and high school students. Viewed otherwise are using only 81% of our capacity. The enrollment is declining. Here are the data from the Facility Plan (actual data through 200102; projections from 200203 on): Here we see that the total enrollment has been declining since 1996. It's hard to know what we are saving 19% of our middle and high school capacity for. Here are the actual capacity data:
The low enrollments also show up as unused space. To start with, here are the middle and high school building areas (square feet) per enrollment of the Richmond schools. Here we see Mosby, Kennedy, and Henderson leading the middle school and high school pack. Attempting to understand the numbers, let's look at the high and middle schools separately. First the high schools: Community and Open are special cases. If we restrict the analysis to the others, we see: The least squares fit is fairly good (R^{2} of 69%) but the intercept leads to the (plainly silly) conclusion that the average common area per kid is 470 square feet. That would suggest that Armstrong, to pick a school near the middle, has 387,000 square feet of common area when, in fact, the entire school occupies only 156,148 square feet. Perhaps more interesting: Kennedy has over 300 ft^{2} per kid while Wythe and Huguenot both have less than half of that. Hang on to the notion of inexplicable and large differences from school to school. Among the middle schools, we see an enhanced example of this problem, with a kid at Mosby enjoying well more than twice the space available per kid at Thompson and Boushall. The least squares fit to these data tells us there is not much to be learned from a least squares fit to these data. The fitted line does emphasize that Mosby and Henderson enjoy much more space than the other schools. Here are these data:
Turning to the number of faculty, the Education Department website gives us a rough baseline: For grades 812 in 200001, the state average is 11.4 teachers per kid. We can't do an exact comparison to Richmond because the Richmond middle schools are grades 68 and the high schools are 912. With that caveat, here are the Richmond data expressed as number of full time teachers per 11.4 students. The first thing that jumps out here is that we have a lot of teachers and the teacher/student ratios vary quite a bit. Separating the high and middle schools, we see a lot of teachers in the select high schools and fewer in the mainstream schools. Leaving out the two select high schools, we see three schools (Armstrong, Kennedy, and Marshall) with more than average numbers of teachers and three others (Huguenot, Jefferson, and Wythe) with fewer. The least squares line gives only a mediocre fit and suggests that the larger schools enjoy some economy of scale (while TJ just has to make do with fewer teachers).
In the middle schools we see an excess of teachers but no particular pattern (except that Chandler gets the short end of the stick): We also have SOL data. These lead to the important question: Do all these extra teachers improve the SOL scores? In the high schools the answer is "no" if we exclude the two schools with select student populations: Indeed, the least squares fit to the data for schools with mainstream student populations says that increasing the number of faculty doesn't help the SOL scores. To the contrary, we see lower SOLs where we have more teachers: The R^{2} says that the correlation with number of teachers explains only about a third of the trend. Even so, the ten percent of extra teachers at Armstrong plainly aren't doing much good. As to the middle schools, there is only a hint of a correlation (<1%) between SOL and number of teachers. Here are the data:
What does all this mean? It suggests that the millions we are spending on extra teachers in elementary and middle and high schools, is wasted money as far as student performance on the SOL test.

Last updated
04/01/12 