Is Shrewsbury’s per-pupil spending really too low ?
by Chris Kirk, February 14, 2014
One of the incessant complaints of Shrewbury’s School Department is that Shrewsbury spends too little to educate its children: its per-pupil spending is just too low. In this article, I’ll provide evidence that it isn’t.
If we compare the per-pupil spending of Shrewsbury with that of towns that the Superintendent of Schools says are “similar” to Shrewsbury, it certainly appears that the officials have a point. Here’s a chart showing the per-pupil spending of Shrewsbury and “similar” communities. (see footnotes at SRT – Per-pupil spending – footnotes – Feb. 12, 2014)
There’s Shrewsbury, second from the right. And on the far left, the town of Berlin.
Now, however, let’s look at the enrollment in each of these communities.
Wow! Suddenly the previous chart seems to have been reversed! Now we see that the town of Berlin – which had high per-pupil spending in 2012 – had a tiny enrollment of about 361 in that year. Furthermore, Berlin’s enrollment is dwarfed by that of Shrewsbury, which had low per-pupil spending but a huge enrollment of 5,947 in 2012 – the largest enrollment of all the communities that are “similar” to Shrewsbury.
(We now also see that the School Superintendent’s definition of what constitutes a community that’s “similar” to Shrewsbury is very broad.)
Since the charts appear to be almost mirror images of each other, I wondered whether there might be some relationship between enrollment and per-pupil spending.
So I created a graph with enrollment along the horizontal axis and per-pupil spending along the vertical axis. I then found the enrollment for a given town along the horizontal axis of the graph and then I went up and plotted a dot representing the per-pupil spending of that town. I then asked a computer to draw a dotted line on the graph which best represented the pattern of per-pupil spending of Shrewsbury and “similar” towns (a “best-fit” line or a “linear regression”, to use the technical term). The result was this graph:
You’ll immediately notice that the dots representing the per-pupil spending of “similar” towns generally lie quite close to the dotted line, which represents the pattern of spending among the “similar” towns. Indeed, most of them – including the red dot that represents Shrewsbury – lie almost on the dotted line itself.
However, what’s even more striking is that the dotted line slopes downward as it advances from left to right. That means that …
as a town’s enrollment increases, its per-pupil spending decreases.
What?! How can that be true?
Actually, this is what you might expect to happen. This is an example of what are called “economies of scale”: Consider an elementary school in Berlin and one in Shrewsbury. Both schools have construction and operating costs (to heat and light the building, to maintain the building and its grounds, etc.); both require a principal and other administrative, teaching, custodial, and cafeteria staff; both require equipment (desks, chairs, books, etc.). However, in Berlin, these costs are spread among a population of only 2,866, whereas in Shrewsbury, those costs are spread among a population of 35,608. Thus, the per-pupil costs of Berlin are much higher than those of Shrewsbury for this reason alone.
Shrewsbury can, and does, educate its children more efficiently than other “similar” towns precisely because it’s so much bigger: it can spread its costs among many more families.
That Shrewsbury’s costs per child are so much lower per pupil is not the result of stinginess on the part of Shrewsbury tax payers; rather, it results because Shrewsbury’s school administrators can take advantage of the efficiencies that are available due to the town’s size.
Instead of denigrating the town’s lower than average per-pupil cost, officials and residents should be proud of it.