Here are some critiques of the study you cited in addition to the critiques already mentioned in this thread. Suffice it to say that it is no surprise the guy you citing cherry picked data to support his preconceived notions.
No, they don’t really. You are misinterpreting the data. One figure is attempting to quantify direct expenditures on education per student while the other is calculating the money spent by DCPS divided by the number of students. The latter is also highly suspect given it’s origin and lack of corroboration. Either way, not all money spent by DCPS is spent educating students, nor are the costs all under DCPS’s control (eg. paying private school fees for students).
You are generally right that vouchers are usually granted for less than the per pupil cost, but your conclusion in misleading. While a crude measurement like per pupil spending may go up, the value each dollar is less. Why?
First, because fixed costs (and most other costs) are not based solely on the number of students enrolled, nor can they be easily changed. So even if I can pawn off a few students to a private school at less that my per pupil spending rate, I can’t reduce my costs by an equivalent amount. I can’t pay my teachers x% less because I lost students. It doesn’t cost any less to heat the school because fewer students are there, and I cannot trade in my building for a smaller one. Losing students might require more money in aggregate because I essentially have the same fixed costs, and now I have to pay the voucher.
Second, even though per pupil costs might go up, it doesn’t tell you how effective the remaining money can be. For example, let’s say my school has 10 kids and I pay $10k/pupil. Let’s also stipulate my voucher is for $7,500. If I lose one kid, the per pupil spending goes up by 2.7% or $277. That sounds great, but that remaining money has to educate students who are usually far more costly to educate. Let’s pretend for a second I can reduce my fixed costs to so that I really have that $10,277 to educate each remaining kid. The problem is that the per pupil spending tells you nothing about how that spending is distributed. While the average kid may cost $10k to educate, some may cost $5k in (more) direct costs, while others cost $25k. A private school taking in kids with voucher of $7.5k will rarely take in a kid that costs more than that to educate them. They certainly won’t take in kids that cost $25k to educate. Thus, you are leaving the expensive kids behind, while the cheap kids leave. That means that even though the average spending per pupil may increase, the average cost to educate each pupil might increase by more.
The combined effects of the two issues I raised, in addition to the increased costs of attending private school borne by the school or the individual voucher recipients are why vouchers are usually not a win-win.