January 22, 2023 Letter

With the Governor’s presentation of his State of the State address on Tuesday and his budget on Thursday, the General Assembly will pick up its pace of work, and I will be reporting to you from my assignments to the Finance, Education and Oversight Committees. In this week’s letter, I will complete my review of the State’s education aid funding formula.

A. The three main components of the funding formula
As noted in my January 1 letter, the formula’s “core instructional amount” (reviewed in the January 8 letter) sets the budget per regular needs pupil. The “student success factor” (reviewed in the January 15 letter) adds money to the “foundation budget” for high-need students. Today’s letter will describe how the State allocates the cost of the “foundation budget” between local communities and the State as a whole.

B. Basic design of the “state share” component
If you wish to learn all of the details of the formula’s “state share” calculation, I recommend pages 3-4 and 6-7 of a Report I prepared in 2016 as a private citizen the last time the General Assembly reviewed the funding formula. My more “reader friendly” description today will not delve into all of these details; instead, I will now present the basic structure of the “state share” calculation, the problems caused by a components of it called the “quadratic mean” and conclude by suggesting possible reforms.
The State has used different types of education aid funding formulas for more than half a century, but the division of cost between the State and local school districts has, throughout this period, followed the same general principles. 

The current funding formula’s calculation of state share begins with what I will call a “baseline State share” that follows these basic principles. First, the State decides how much of the total school budget will be paid with State funds (usually around 50%, currently 52.5%). Second, that global amount is distributed among local school districts based on each community’s ability to pay, as measured by the size of its property tax base per pupil. If a community had the State average of property per student, the State will pay 52.5% of that community’s “foundation budget.” If the community has less than the State average tax base per student, the State share will be greater, exceeding 80% in urban core communities. On the other hand, if a community has a wealthy tax base, then the State share will be less than 52.5%; indeed, if a community’s tax base per student is especially rich (as in more than 2.1 times the State average), the “baseline State share” is zero, because that community can pay its entire school budget at a lower property tax rate than the State average.

If the funding formula ended with this “baseline” calculation, it would follow consistent past practice in our State, common practice in other states and be eminently fair in my opinion. Unfortunately, Rhode Island’s formula takes another step that causes it to deviate from these principles by incorporating the “quadratic mean.”

C. The “quadratic mean” and its anomalies
To calculate the “quadratic mean,” the current funding formula adds another factor, namely the proportion of children in poverty residing in a particular school district. It combines these elements in this algebraic formula:

The “quadratic mean” is unique to Rhode Island, and is needlessly obscure, quite possibly by design. To begin with, one should note that the proportion of children in poverty living in a school district already is built into the “student success factor” of the funding formula, and the extra needs of communities serving those children are already accounted for in that way. It does not need to be counted a second time in determining “state share.” Why, then, did the State enact this redundant and/or counterproductive complexity? I believe the “quadratic mean” served a political goal of providing a certain level of State funding to every community with children in poverty, regardless of how much property wealth that community has, or to put it another way, how low the property tax rate in that community happens to be. In my 2016 report, I provided two charts to illustrate the anomalous results of this factor.

1.Transferring State aid from poor communities to wealthy ones

Chart 1 below illustrates how the “quadratic mean” transfers State aid from poor communities to wealthy ones. The left column in the first block lists the seven communities with the largest tax base per student, all of which have more than double the State average. As a result, their “baseline” state share (described here as “adjusted EWAV” is zero. The quadratic means transfers to these communities State aid for costs that they are relatively able to pay on their own, in a total amount of $22.6 million.

The left column in the second block lists four of the State’s communities with the smallest tax base per student. For three of them, the “quadratic mean” reduces their State aid by between $2 million and $3 million, effectively transferring a total of $8 million from the State’s poorest communities to the wealthiest. (Providence does gain State aid through the quadratic mean, which it may lose if this single factor is adjusted in isolation; however, I believe Providence’s overall aid would increase if the funding formula were subject to a general reform that addressed most or all of the issues in these letters.)

2.Providing disproportionate aid per student in poverty

The stated rationale for the “quadratic mean” is to provide extra help to communities with children in poverty. In Chart 2, I calculated how much State aid the “quadratic mean” provides above and beyond the “baseline” State share for each child in poverty in the same seven wealthy school districs and four poor school districts as in Chart 1. The calculations are as follows:

Looking to the rightmost column, the calculations reveal that in 2016 the “quadratic mean” rewarded the State’s seven wealthiest communities (whose property tax wealth is more than twice the State average), increasing their State aid per child in poverty by an average of $7,308, while reducing State aid per child in poverty by $339 in Pawtucket, $1,510 in West Warwick and $566 in Woonsocket.

D. Conclusion

I was not “in the room where it happened” when the “quadratic mean” was added to the funding formula in 2010. If the political goal of this element was to gain support from wealthy communities by providing State aid beyond what a reasonable funding formula would require, I believe it would be more transparent to remove the “quadratic mean” and add a minimum state share of 5% or 10%. In this way, we could make the “state share” component of the funding formula easier to understand and also more transparent and honest in its operation as a real funding formula that works in the way it claims to.