This post is the second in a series (of unknown number) focusing on how states harm local public school districts through illogical, ill-conceived state school finance systems and components of those systems. One goal of this post is to illustrate the types of problems/manipulations that exist in state school finance systems, how they work, and the severity of the problems they can cause.  I have written previously, for example, how states find ways to actually use state aid to make their finance systems less equitable (school finance pork). I have also written about policies like census based financing of special education and it’s adverse effects on high need districts. The Commonwealth Triple-Screw takes it to another level.

The Commonwealth of Pennsylvania has among the least equitable state school finance systems in the country. Pennsylvania operates a school funding system that on average provides systematically less state and local revenue per pupil to the state’s highest need large and mid-size city districts. Among the nation’s most “screwed” city districts are Philadelphia, Reading and Allentown.

But amazingly, in Pennsylvania, the pain doesn’t end there. Pennsylvania also has one of the least fair, least logical approaches to special education funding, both in terms of the way in which special education aid is distributed to local public school districts and in the calculations for determining how much should be paid by local public school districts to charter schools for serving special education students.

Apparently, this issue is of current interest in PA: http://www.mcall.com/news/local/parkland/mc-lehigh-valley-cyber-charter-schools-20120604,0,2970776.story

The hit comes in three parts and I call it the Commonwealth Triple-Screw. Here’s a run-down.

Screw 1: Census based financing of special education, assuming a uniform share of students in need on its face provides less support to districts with greater shares of students in need.

First, Pennsylvania is among a handful of states which continue to use an approach called Census Based financing of special education. In brief, what PA does is provide to each school district, in general regardless of their local wealth and regardless of their actual number of special education students, a flat, base allocation of special education funding per 16% of total enrollment (the assumed flat % special ed across districts).

The argument is that funding special ed in flat amounts avoids the incentive to over-classify students. This argument ignores the possibility – the simple reality – that populations of all types vary in their geographic distribution for a variety of reasons and that includes families of children with disabilities. Funding on this basis necessarily deprives districts that through no fault of their own have far more than 16% special education. Further, and more illogically, for districts having only about 7% special education, this approach arbitrarily over-funds their needs (at least relative to higher need districts). I have written extensively about the research and realities of Census Based funding in this recent article:

• Baker, B.D., Ramsey, M.J. (2010) What we don’t know can’t hurt us? Evaluating the equity consequences of the assumption of uniform distribution of needs in Census Based special education funding. Journal of Education Finance 35 (3) 245-275

Here are a few quick snapshots of how this works out globally (Statewide) then locally (Chester Upland School District). The following figure is drawn from my Summer 2011 final update to my testimony in the case of C.G. v. Commonwealth (a Federal Court challenge to PA special education funding). The figure shows that districts with higher special education population shares and higher Market Value/Personal Income ratios (hence low wealth/income) generally receive less special education aid per special education pupil as a function of the underlying census based formula.

Figure 1. Special Education State Aid per Special Education Pupil

In the rest of this post, I will show in particular how this formula along with other calculations dramatically undercut the financial viability of Chester Upland School District, a high need district facing dire financial circumstances in recent years.

Table 1 provides a walk-through of Chester Upland’s position with respect to special education state funding. CUSD was reported to have been allocated just over $5 million in SEF funding for 2010-11, with that funding having been frozen for the past several years. CUSD’s actual percent of enrollment in special education programs has typically been about 22% over time (based on several sources, including the NCES common core of data). That would amount to over 1,500 special education students, consistent with counts reported later. This yields special education funding per actual special education pupil of about $3,200(placing CUSD among the red squares in Figure 1 which lie at approximately 22% special education and have low wealth [high MVPI]).

Table 1 walks through the hypothetical difference in funding that would occur if CUSD was allocated SEF per actual pupil rather than per 16% ADM. CUSD’s SEF allocation per 16% ADM is $4,429, or $1,200 per pupil higher than it is per actual pupil in need. If CUSD received $4,429 per actual child in need, then CUSD would receive nearly $2 million more in special education funding, or a 38% increase. For a district with minimal local capacity to offset this loss, that’s a significant hit. But, it’s also the smallest hit of the triple-screw!

Table 1. Special Education Funding in Chester Upland

[a] SEF data from: http://www.portal.state.pa.us/portal/server.pt?open=514&objID=509062&mode=2

[b] MVPI Aid Ratios from:http://www.portal.state.pa.us/portal/server.pt/community/financial_data_elements/7672

[c] IEP % Data from: http://www.nces.ed.gov/ccd/bat

Screws 2 & 3: The charter school funding formula for special education students exacerbates these problems by requiring high need, under-resourced districts to pay special education tuition to charter schools at the rate of the average special education expenditure per special education child of the host district.

Pennsylvania’s formula for determining the amount of money that must be transferred from sending districts to charter schools for serving students with disabilities is poorly conceived, creates perverse incentives for charter school operators, and inappropriately drains disproportionate resources from sending districts. CUSD is perhaps more harmed by this ill-conceived mechanism than any other district in the state both because of the characteristics of CUSD and because of the enrollment practices of Chester Community Charter School.

Note that the following analyses present the hypothetical adverse effects of the Pennsylvania charter school funding formula on CUSD, using enrollments of Chester Community Charter School to illustrate those effects. CUSD may also be sending children with disabilities to other charter schools, where the effects would play out similarly to the extent that other charters also siphon off children having lower cost disabilities. Further, the hypotheticals are based on enrollment-by-classification data from 2008-09 and conditions may have become even more severe in recent years.

The funding hit comes in two parts. First, Table 2 shows how the spending special education tuition rate is set. CUSD spent in 2011-12 about $17.3 million in “selected” special education spending. Even though this spending was used for approximately 22% of the district population, the tuition sending rate is calculated per only 16%, incorrectly inflating the special education spending per special education pupil.  16% of CUSD ADM is just under 1,200 students. Thus, the special education expenditure divided by that figure is $14,670. This is the “additional expenditure” per special education child. Each special education child also has associated with him/her a “base” (regular education) expenditure. That figure for CUSD is $9,858. Therefore, the total including BASE + SE is $24,528. If we take the $24,528 and multiply it times the total number of special education students sent to Chester Community Charter School, that amounts to over $15 million.

Okay… so let’s stop for a minute. The district receives in SEF funding about $3,200 per special education student, and must send out over $24,000. That’s difficult enough… but… the $24,000 is miscalculated substantially in two ways!

If we consider that CUSD actually served about 22% special education (or about 1,620 students),the special education spending per pupil, with the base added in, would be $20,527. If we use this figure instead to determine the payment to the charter school, CUSD would send only $12.7 million to the charter.

This arbitrary use of the 16% figure to determine sending tuition rates costs CUSD nearly $2.5 million (or about 50% of its state special education funding)!!!!!!!

Table 2: Over-expenditure to Charter Special Education Part I

But this is only the first and smaller portion of the miscalculation of sending tuition rates for special education students. Table 3 shows distribution of counts of students by disability type for CUSD and for Chester Community Charter School. Also in Table 3 are the average additional expenditures per special education student based on analyses from the Special Education Expenditures Project (SEEP). I’ve chosen to use these average additional expenditure margins so that I can simulate the effects on host districts of charter schools choosing to serve only the least needy (and least costly) special education students. 92% of the children with disabilities in the charter school are those with the lowest cost disabilities, compared to only 66% in CUSD. Yet, CUSD must pay out to the charter on the basis of an already inflated average special education cost per special education pupil.

Table 3: Distribution of District & Charter Special Needs Students & Related Cost Margins

Table 4 provides a walkthrough of the estimated impact of this financial hit. First, if we take the actual special education spending per actual special education pupil in CUSD ($10,699) and express it with respect to the spending per average non-special education child in the district we get a ratio consistent with research literature over time. On average, in CUSD, the special education child is allocated just over 2.0 times the average non-special education child (where the base non-SE child is just under $10k and the special education spending margin is just over $10k, see Table 2). I have performed this check simply to see that average special education spending margins in CUSD are in line with prior research findings.

In Table 4, I estimate the additional expenditure of an SLD child by taking the research based additional expenditure ratio of 1.6 and multiplying it times the average additional expenditure of a non-special education child in CUSD (1.6 x $9,858).  That gives me an estimated additional expenditure per SLD child of $15,774. I do the same for children with speech impairment, yielding an estimated additional expenditure of $16,759. Indeed, even the charter serves some children with disabilities estimated to have higher than average additional expenditure, but very few of them.

Calculating the charter payment based on the inflated district average (based on the error mentioned in Table 3 above), CUSD must allocate over $15 million for children with disabilities in Chester Community Charter School.

But, if we re-calculate the charter allocation using the additional expenditure ratios from research, where some children will have higher than average cost and some lower than average cost, we find that CUSD would need to allocate just over $10 million.

In short, the district is overcharged by nearly 50%. The district is overcharged by an amount equal to nearly all of the district’s state special education funding, even though the district is left with more than half of the total special education children to serve and nearly all of those with more severe disabilities.

Table 4: Over-expenditure to Charter Special Education Part II

Arguably, this mechanism actually provides incentive for Pennsylvania charter schools to seek out, recruit and serve children with mild disabilities, creating similar budget pressures for other districts across the state.

Cumulative effects of the Commonwealth Triple Screw

Chester Upland School District’s expenditure budget for its own students is now approximately $54 million (after transfers to charters).  Note that if Chester Upland had received special education revenue from the state based on actual percent special education, the district would have received about $2 million more in revenue to spend, some of which would have been transferred to charters for serving special education kids. But let’s assume that about half should have stayed with the district. So, there’s a $1 million hit to start. Then there’s the big double hit, which amounts to $4.8 million!

So, we’re talking about a cumulative hit of, oh… hypothetically… about $5.8 million, or over 10% of the district’s budgeted expenditures (in other words, they should have received and kept roughly an additional million in state special education funding and should pay out a simulated/estimated $4.8 million less to charters for special education students).

And that, my friends, colleagues, co-bloggers, tweeters and avid readers is the Commonwealth Triple-Screw!