School Funding Deception Alert! (in a CAN)

I’ve noticed a pattern in a few recent school funding proposals, mostly emanating from shoddy, haphazard proposals developed on behalf of the CANs (ConnCAN & its close relatives) and often with “technical support” of Bryan Hassel of Public Impact. Let’s call it school finance reform in a CAN.

These new simplified school funding formula proposals, framed under the “money follows the child” ideology are intended to make state school funding formulas more “transparent” and to allow for more equitable and predictable flow of funding to charter schools or other non-district schools.

In each proposal (ConnCAN’s Spend Smart & The Tab, or Rhode Island’s new formula [albeit laced with other problems unique to RI-see post]), among a variety of other major overlooked factors, arbitrary and unfounded recommendations, exists a seemingly innocuous proposal regarding how to target aid for variations in student needs across districts.

As the authors of ConnCan’s recent Spend Smart brief explain deeply embedded in a footnote… you really only need to use a single factor to get state aid targeted to the right schools and that factor is the share of children qualifying for FREE OR REDUCED PRICED LUNCH. There’s no need for a special factor for limited English proficient/English language learner populations, or anything else. It’s all pretty much correlated to free and reduced lunch. (Hassel’s previous report for ConnCan, The Tab, included a trivially small LEP/ELL weight instead of none at all).

First, this assumption is patently wrong to begin with and is never actually validated by the authors of these proposals. But let’s set that aside for the moment. I’ll have a future post where I use actual data to show just how freakin’ wrong the assumption is.

But why would they propose this anyway? Well, it turns out to be really simple. If a state has a fixed sum of money to distribute (generally how it works), the CAN game is to figure out on what basis might charter schools get the maximum share of that money – regardless of who really needs it most. That is, what measures CAN they choose for weightings which will drive money to charters. Charter schools do tend to operate in poorer communities (relative to state averages), but a) serve the less poor among the poor, b) serve few or no LEP/ELL children, and c) incidentally, also serve few or no children with disabilities (as has been addressed on my blog regarding NY and NJ charter schools, and will be addressed soon regarding CT charters – numbers already run, charts forthcoming).  I’ll set aside c) for now.

So, the way to maximize charter funding, is to give a single weight for children qualified for free OR REDUCED PRICE LUNCH, and to negate any weight for LEP/ELL children (or make it as small as possible). That way, charters will get the same weight for kids whose family income falls between the 130% poverty level and 185% poverty level as neighborhood schools get for children below the 130% poverty level (This distinction is NOT TRIVIAL), where neighborhood schools have far more of the lower-income children. Any money that would have gone to LEP/ELL children can be rolled into a bigger weight for free/reduced lunch children, channeling a larger share of the total funding available to charter schools.

While not specifically addressed in these proposals, one would imagine that the same pundits would also favor flat, lump sum, or census based funding for special education, not differentiated by disability type, such that every school or district gets a specific dollar amount for special education based on a fixed share of their enrollment – a) whether they serve any special education students at all, or b) whether they only serve mild specific learning disability students, and none more severe. Watch out for this one as well!


Note: I’m sure that many will respond to this post by arguing that charters get severely shortchanged on state aid anyway, and that even if they make out okay on these adjustments, the lack of funding for such things as capital outlay and facilities more than offsets the difference. That’s a topic for another day. But, suffice it to say, existing comparisons like those made in the recent Ball State/Public Impact (imagine that) study are grossly oversimplified (as I explain regarding NYC schools, here: (page 23)). For example, typical crude comparisons never address whether having few or no special education children (relative to averages for district schools) result in cost reductions (per pupil) that might actually be greater than facilities average annual expenses per pupil.


Follow up figure for those who asked:

Note that using only a weight on free or reduced lunch would drive the same amount of supplemental funding to Torrington as to Norwalk or Danbury, despite large differences in LEP/ELL populations. So too would any charter school that has comparable low income population to Danbury, Stamford or Norwalk, even if that school had no LEP/ELL children. There may be other valid differences that require additional attention. Even this graph is too crude to give us the full story. The bottom line is that one must at least evaluate the distributions of children by need categories across districts and settings before making such bold, but oversimplified policy recommendations.

Here’s Rhode Island:

The issue here is similar in that Central Falls in particular (imagine that) gets shafted by failure to independently address differences in ELL/LEP concentration. While RI has few districts, and has a specific cluster of high poverty districts, the rates of LEP/ELL children across those districts vary from 5% to 20%. But, as I’ve explained previously, the RI formula and logic behind it have numerous other empirical and logical gaps. see:



  1. I’d be careful before just lumping different states together on this. Different states have dramatically different enrollment patterns and different charter sectors. Rhode Island, for example, sees very little of the effects that you’re observing in NJ (there are typically much more free lunch students in charters than reduced price lunch students as is true in the districts they draw from). And the ELL/LEP populations are very tightly packed into just several high poverty districts in RI, so while not having an ELL/LEP weight is against the student-based funding principal in its most exact form, taking the perspective of, the end result matters.

    Additionally, RI had been clear that they were starting with something more simple and transparent to just get a formula in place to begin with and start to tweak after reexamining financial data a few years out. Any tweak that sends substantially more money to the poorest communities and the communities that have the most ELLs is a pretty good start, and adding weights and additional considerations is a lot easier from a policy perspective than removing them later. Their new accounting work in RI should lead to a far deeper analysis of school revenues and expenditures that may have a substantial impact on the continual updating of the funding formula.

    My understanding of ConnCAN’s current proposal is that they view the situation in CT as only being slightly better than RI. There is a formula, though it’s been so manipulated and played with it hardly factors into decisions, but it’s very complex and still uses the ten year census data on poverty as a major element as opposed to a more updated data source. They’ve been pushing to basically start fresh with something that’s far more simple and uses data sources that can be updated annually. I’m not sure what the situation in CT is with free versus reduced price lunch.

  2. I’ve noticed in some urban settings there is about 100% free and reduced in the elementary grades. Once the child goes to high school that number drops off significantly. Do the kids and families get richer over the summer between 8th and 9th grade? I think the kids don’t allow the forms to get filed to avoid the “stigma” of having to get free and reduced lunch. I wonder if it is still true or true everywhere there are teens trying to look cool.

Comments are closed.