Alert reader davidc (THANKS!) pointed out that RIDE has posted backup documents to a new education funding formula page here, and I'm happy to report that we now have enough information to calculate the state share (I've updated the GoogleDocs spreadsheet with this info, and will be using references to that worksheet as I go through the explanation below.

But first, a couple of comments. Although I had sent an e-mail to Deputy Commissioner David Abbott (cc'ing Commissioner GIst) asking for the information, they did not have the courtesy of replying to let me know this had been posted. And while I don't know when the page went live, forensic examination of the supporting data PDF shows that it was created Thursday morning. (Line 1513 for those geeks playing along at home: /CreationDate(D:20100304155512Z)/Author(Beardsley, Andrew)/Creator(MicrosoftÆ Office Word 2007)/Producer(MicrosoftÆ Office Word 2007)/ModDate(D:20100304113952-05'00')

Second, I can't help but point out the glaring error in their document. On page 2, the total adjusted assessed value state-wide is this:

But on the following page, this is the number given:

Now I know that a single digit off on a hundred-billion-dollar number is insignificant. But if my son handed that in to his math teacher, Mrs. Warner, she'd circle it, and with good reason. If you want people to trust your math, getting the little things right — like having the same number on two successive pages — is kind of important. I mean, really. What would Arne Duncan say.

Third, and this is a subtle methodological point, but I question the amount of rounding that occurs in the RIDE calculations. Because both the percentage of students on Free and Reduced-Price Lunch (FRPL) and the Student Share Ratio for the Community (SSRC) get squared in the formula, small changes get magnified. They also appear to have rounded the final State Share number. I've included columns in the spreadsheet indicating the differences between calculated values and RIDE's rounding. It's mostly in the tenths-of-a-percent, but when we're talking about allocating taxpayer dollars, it doesn't cost RIDE any extra to use a couple of decimal places.

At a high level, the formula is fairly straightforward. RIDE's consultant at Brown developed a "core instructional amount" based on an average of four New England states (RI, MA, CT, and NH). This value, $8,295, includes instructional costs designated by the Basic Education Program (BEP). It does NOT, however, represent the cost of actually educating students, since RIDE acknowledges that it does not include teacher retirement, group home aid, food service, transportation, safety, building upkeep, utilities, maintenance, debt service, capital projects, out of district tuition and transportation, and non-public textbooks.

In other words, this number represents only some theoretical ideal, not the actual cost of opening the doors of a school building. Let's just be very clear on that.

Now, let's see how much of this number RIDE considers the state should be providing. To get there, we need to multiply $8,295 by the number of students in the district, which is referred to as the Resident Average Daily Membership (RADM). To that number, we need to add an adjustment for the number of students in the district who are receiving Free or Reduced-Price Lunch (FRPL) which RIDE is using as a proxy for the percentage of the population who are economically challenged. RIDE assumes that these students will need extra help, so they add a factor of 40% of the core $8295 for each FRPL student.

So the total amount the district "should" be spending is RADM * 8,295 + (FRPL * (8,295 *.4))

If we put in the numbers for Portsmouth, with a RADM of 2,657, we get a result of $22,879,269. The obvious question — how to reconcile this with our current school budget of $35,746,286 — does not seem to enter into RIDE's calculations. They are not talking about the cost of putting the teacher in front of a room that has heat, light, and buses to get the kid there. But wait, it gets better.

The heart of the state funding formula is a quadratic mean, which weights the community's ability to pay (the State Share Ratio for the Community, or SSRC) and the percentage of students on FRPL.

To get the value to plug into SSRC, we need to do some further math with a number called the Equalized Weighted Assessed Valuation, or EWAV, divided by the number of students in the district.

So for Portsmouth, SSRC = 1-(0.5 * ($4,549,808,726/2,657)/(141,599,970,284/139,934) or 15.39% (That's cell S40 in the spreadsheet, RIDE rounds this.) Now, we plug that into the quadratic mean. Since Portsmouth has 1,270 K-6 students of which 122 receive FRPL, our percentage is 9.61 (another one of the things that RIDE rounds).

Therefore, state share equals the square root of (15.39^2 + 9.61^2 / 2) or 12.98% (cell U40).

Now, we can use that percentage, with the "core" cost and we'll find that the state would contribute $2,968,986. I won't make you do the math: that means local taxpayers need to pony up $32,777,300.

So you take a hypothetical number (which acknowledges that it does not represent the full cost of educating a child) and then jump through a series of mathematical hoops, rounding here and there, and come up with a percentage that bears no relationship with the actual per-pupil costs in a district. I hope that RIDE can forgive me if I seem unconvinced of the "fairness" of this.

Note: A reader suggested calculating the percentage of the total cost the state would be picking up: 9.06%.