Correlation does not imply causation

Below, in black, are the talking points issued by the Department of Education last week to pressure legislators to shift their support away from the Damon amendment to LD 1932 which expands the school reorganization law to allow reorganizing districts to adopt an alternate governance structure of regional school unions.

Senators, who had already voted to include the amendment in the bill, were told that by Department officials that every student in a new regional school union would represent $1000 that could be unavailable for their own districts.

Our own annotations to the talking points are in red (Brian Hubbell) and blue (Judy Sproule).

MDIschools.net, 2/17/2008

School Unions, Efficiency and Performance

· Maine spends $24 million more per year than necessary on the education of 26,719 students educated in school unions, and with no added academic benefit.  That’s because school unions as a group spend more per student than in other types of school units with no better test results.

By this logic, one could also assert that every dollar spent in any school system above the per-pupil average is “unnecessary”.  What is “necessary”? If they are using EPS they are misguided. Using average cost as a benchmark also lacks merit. Let’s have a real definition of necessary, and then it is up to the municipalities to determine what they want to pay for that is additional, and none of the state’s concern – it’s not their money.

· 62% of the 50 most expensive K-8 school administrative units (SAUs) in the state, regardless of size, are in school unions. And 38% are not. Look at what is responsible for the costs. Maybe it is a remote location; maybe the taxpayers want to include foreign language instruction, etc. Only 28% of the least expensive K-8 SAUs in the state, regardless of size, are in school unions. Are they looking at unions that have SADs and CSDs as members? There are more ways to slice this bologna than one can count.

This is the logical fallacy that correlation is proof of causation.  David Silvernail’s analysis in fact shows that the variable with the greatest correlation to a school system’s per-pupil expense is municipal valuation.  Wealthier towns spend more on education, less wealthy towns spend less.  The statistical correlation that the Department has cherry-picked above may indicate that wealthy towns have found some cooperative advantages in union systems.  Just as significantly, it shows that union structures represent the full range of per-pupil costs. This is Maine. A lot of things cost more here. There are several Maines within Maine and, depending on where you are, the costs will be higher.

· For elementary school units, school unions averaged $8,537 per student – $1,009 more per student than in consolidated units (SADs and CSDs).  For K-12 units, school union expenditures are $1,385 (18%) more per student than in consolidated units.

So, is the Department asserting that these same schools with the same programs of instruction could be operated for $1000 per pupil less under a different structure of governance? Yes or no. If the answer is yes we need to ask for a concrete illustration, and I mean one that is financially accurate, not at the level of incompleteness and inaccuracy provided so far by DoE.

· More than two thirds of the school units with 500 or fewer pupils are in school unions. These small units accounted for only 25,675 (12.8%) of the state’s students in 2005 and their average per pupil operating expenses were higher than in larger units and than the statewide average. ($8,990 per student, compared to $7,632 for units with 2,501 to 3,000 students, and $8,230 statewide average).

Small schools in areas of low population density may never be able to match per-pupil costs in dense suburban areas.  But unions are vehicles for small schools to share administrative expenses, not duplicate them.  Unions allow small schools to continue to benefit rural communities where they otherwise, under state or regional authority, might be closed.  This highlights the absurdity that unions are categorically less efficient than stand-alone municipal units. Unions allow municipal school departments to achieve more efficiency, not less. Of course there are economies of scale in larger populated areas.

· In any structure that allows a Regional School Unit or regional union board to delegate to local school committees the authority to hire, fire, negotiate, or raise funds, the costs will be higher.  The higher costs in unions are part of the basis of EPS calculations and thus also raise the statewide cost of education –explain this one! What about those 1800 elected school board members who are theoretically working for free? Reduce the number of volunteers, increase the paid additional administrative bureaucrats that will have to be added, and then add up the additional costs.

o Because a local municipal school committee has the authority to raise and administer funds, as well as hire and fire, and negotiate collective bargaining it must be viewed both legally and practically as a separate school administrative unit. Many would be happy with this conclusion, but it remains only the Department’s own conclusion and objection.  The Damon amendment defines the centralized Regional School Union as the school administrative unit.

o In a union system, citizens cannot see or assess a comprehensive K-12 budget and make educated decisions about prioritizing spending in their district, which also makes finding savings more difficult. I don’t think that is true. The citizens who develop the budget have a far clearer view than a remote board. The person who wrote this also probably never attended a town meeting.

o It is unclear who is responsible for a child’s education – the “super union” board or the local school committee. To whom does a parent turn when their child’s educational needs are not met? The local school board is responsible, and that is the way it should be.

o Decision-making is proliferated among multiple boards which makes impossible clear, consistent K-12 curriculum, budgeting, priorities, etc. That is completely untrue.

This is a collection of imaginary foibles that disturb only the Department’s dreams.  As with the Department’s Regional School Units, the boards of Regional School Unions would be comprised of the same representatives who serve on the local boards. Citizens dial their same representative’s telephone number either way and the same representative has oversight over both regional and local concerns.

Figures above include all districts for which 2005-06 data was available.

In summary, the school union structure is Maine’s highest cost form

of school governance with only mixed results on performance.

• Too many layers of administration.  In most school unions, a single superintendent and central office reports to multiple school boards, develops and administers multiple budgets and, in many cases, negotiates multiple collective bargaining agreements.  Superintendents in these systems have often complained they spend too much time reporting and don’t have the time they need to focus on educational programs. In some unions, there are multiple central offices, adding to the duplication of functions.

Superintendents also find productive connections with their schools through contact with local boards and municipalities.  Local boards allow efficient division of effort and specialized expertise. The Damon amendment centralizes regional administration and allows common bargaining agreements. The reporting burden is created by the DoE.

• Lack of K-12 alignment.  In most school unions there is a lack of an aligned K-12 curriculum.  When multiple school boards are deciding curriculum and offerings for elementary school programs, students from those schools do not all come to the high school with the same background?  Are we positioning our children for success or failure?  Can all entering freshman access a rigorous curriculum or will they have to settle for less because they didn’t have comparable experiences?  The data demonstrates that math, language arts and science curriculum vary within unions. Our children deserve a cohesive equitable education.

Under the Damon amendment, Regional School Unions have fully coordinated common curricula. Unions can have a coordinated curriculum – nothing precludes that. Also, some unions have students going to more than one high school. They need to be equally prepared for all options, not just channeled into one track.

• More money, mixed results.  As a group, school unions are more expensive than other governance structures.  Not only does this cost local taxpayers more, it contributes to the statewide costs which are born by taxpayers in all communities around the state.  That’s because EPS allocations are based in part on actual spending.

The connection between governance structure and higher average cost is a logical fallacy.  School policy and spending is set by local tax payers through local democratic process. Areas of high valuation already receive minimum subsidy for education from the state and so, comprehensively, the state’s general fund benefits more from these areas than it is hurt by them.

This attack on unions is strictly a political ploy to force the attackers’ preference for legislation. Unions can be just as efficient as any other structure by collaborating with other entities, and it does not require changing governance.

For those school units where it made sense for them to form SAD's, they have already done so - and historically there were incentives. For those school units that exist today that are not SAD's, the disparities in valuations and student population are such that it clearly does not make sense for them to "consolidate". This current law is an attempt to force what the logical, free market dynamics do not embrace.

Now, look at these towns which are holding out as SAU-union members. For the most part, the school budget is virtually all of the town budget (in half of the Union 92 towns, the school budget including the state share is more than the total town budget). If the state wants to coerce these towns to turn over their schools to a new quasi-governmental entity, they are really demanding that they abdicate their sovereignty, because once you take out the school factor, what is left? Solid waste removal? Incidentally, that new quasi-governmental entity is conveniently subject to control by the Department of Education. Why would anyone agree to do this?