Cost Per Pupil : Facts, Not Spin

Guest contribution by Arthur Green

Timberlane has long and loudly declared that Timberlane’s per pupil cost is below the state average.  The implication is that Timberlane is a sound financial manager, pinching pennies and making do on less money than other districts.

To compare ourselves to the state average is self-serving and misleading. And with the most recent number just released by the Department of Education, Timberlane is now well above the state average in Cost per Pupil (CPP).

Timberlane’s Cost Per Pupil  (CPP) for the 2013/14 school year ended in June shot up 8.8% from the previous year – almost $1,200 per student.  Both the percentage increase and the absolute number are well above the state average.  (See chart below)

If the state average CPP is not a suitable comparison, what is?

The state has over 160 school districts, many of them small and rural, with average enrollment of about 1150.  The average district in New Hampshire has much different challenges and cost drivers from Timberlane.  To say, as Mr. Collins does, that over 100 districts have higher cost per pupil is a meaningless and misleading comparison.

There are only 9 school districts which have a composition similar to Timberlane

  •  Enrollment in the range of 3,000 to 5,000
  •  Service all grades K – 12
  •  Southern NH
  •  Single HS, MS, cluster of feeder elementary schools all administered as a single district

These comparable districts are Bedford, Concord, Dover, Hudson, Keene, Londonderry, Merrimack, Rochester and Salem. To make the comparison more useful, I will focus just on those districts which have stronger NECAP results in both Grade 11 Reading and Math.

Here is the CPP  for just these 5 “Leading Districts,” with comparison to the state average and to Timberlane:

The Cost Per Pupil average for the “Leading Districts” is well below the state average. It makes sense that districts in the 3,000 to 5,000 enrollment range have more opportunity for economies of scale. In this group, only Keene is higher than Timberlane. Comparing to the cohort of “Leading Districts,” Timberlane’s CPP is significantly higher, and grew faster. CPP for the “Leading” cohort increased 4.8% last year, but Timberlane increased 8.8%. No district in the “Leading” cohort had as large an increase as Timberlane.

And keep in mind, all five of these districts deliver stronger academic results.

What can we say about Cost Per Pupil for the current year 2014/15?

The CCP can’t be known with certainty until the end of each year, but it is possible to estimate it based on our best current knowledge of the budget, spending and enrollment.

I’ve estimated the CPP for this current year ending in June 2015 at $15,484.

I have been criticized for making estimates, but I consider it a duty of people in office to do their best to understand the consequences of their actions and to to tell people honestly about them.

Here’s the math.

Cost Per Pupil has a numerator and a denominator.

The numerator is cost.  Not all costs are included in the DOE calculation, but the percentage increase should provide a good estimate.

Actual costs in 2013/14:  $63,689,780

Forecast costs for 2014/15:  $65,436,000 (based on the budget of $67,736,000, minus the district forecast of a $1.9 million surplus passing forward to 2015/16.)

Percentage increase: 2.7%

The denominator is number of students. For the Cost ­Per ­Pupil calculation, the DOE uses the Average Daily Membership (ADM), which is based on recording actual daily attendance through the year. However, the annual changes in ADM track very closely with changes in the enrollment figure, which is reported as of Oct. 1 each year.

Actual enrollment Oct. 1 2013: 3,922

Actual enrollment Oct. 1 2014: 3,773

Percentage change: ­- 3.8%

Putting it together we have:

Cost: ($14,498 x 1.027 [2.7% increase]) Divided by Number of Students:  (1 – 0.038 [3.8% decrease]) = $15,484 per pupil

This would be an increase of 6.8% ­ – almost $1,000 per pupil – from the year ended in June 2014.

And this estimate is very generous to the administration, because it incorporates the forecast of $1.9 million surplus. We know that the Board and the administration are already planning to make expenditures which are likely to significantly reduce the surplus, increasing this year’s CPP.

Cost Per Pupil is going up dramatically because we are losing students rapidly while allowing our budget to grow.  Do you think paying an estimated $15,484 for every student from kindergarten to 12th grade is value for money when other comparable districts our size with better academics are doing it for less?

Responsible Decision Making

The Cost Per Pupil discussion highlights a key difficulty of making reasonable judgments about the school budget – lack of up-to-date information.

The 2013/14 school year ended last June.  The final CPP number was published by the NH Department of Education on Dec. 15.  Between those dates:

  • Timberlane published as District Report Card on Aug. 7, reporting the old 2012/13 number – $13,329 per pupil – as the most current available.
  • The SAU submitted its “DOE-25” filing to the state on Aug. 29, reporting a preliminary CPP figure of $14,414
  • The fall budget season began.  The complete Draft 1 budget was published on Nov.17.  By Dec 11, Draft 3 was being reviewed.  In all the deliberation to this point, the only “official” CPP number acknowledged by the school district was the 2012/13 number, $13,329

For those who prefer to avoid acknowledging the rapidly rising costs in the district, it is very convenient to have a rear-view mirror which only shows the data once actual events are 18 months in the past.

My view is that CPP is an example of an important financial metric which needs to be understood by everyone involved in the budgeting process, and which needs to be actively used as a tool (one amongst many) to gauge spending plans.




Filed under Budget 2015-2016, Budget Committee, Expenditures, The Mushroom Farm

28 responses to “Cost Per Pupil : Facts, Not Spin

  1. Kat

    When you say other districts have better academics, do you mean that they have better NECAPs?

    • 11th grade math and reading NECAPS. This is a standardized test across the state and a reflection of a school district’s “total effort” so to speak. No one has proposed any other measure of academic excellence. Mr. Collins suggested AMO but that is a measure of progress not accomplishment which my blog has debunked in the Dec. 7th posting.

  2. Kat

    That seems like a very limited lens through which to view and measure any school district. I’ll try to collect my thoughts a bit more and elaborate.

    • I understand your point but the problem is that academic achievement has to have a measure yet there are few measures that are universally available. It is important that districts are not universes onto themselves. They must stand in comparison to each other for the benefit of students, parents and taxpayers. Will look forward to your further thoughts and thanks for writing.

  3. Bustmore Gas

    Just more Green fuzzy math.

  4. Concerned Parent

    Out of curiosity, why is Southern NH used as a comparison? I think there’s more than enough criteria to make a good comparison without limiting the scope to So. NH (which is defined as what? South of Concord? South of Manchester?).

    What do the numbers look like without that criteria?

    • The primary consideration with comparable was student population and that all districts have to operate their own schools from K to 12. The biggest districts are in So. NH.

      I would be very interested in another way of selecting comparables if you want to try. All info is on the DOE site.

    • Arthur Green

      Thanks for your thoughtful question. I didn’t exclude any district in the 3,000 to 5,000 student range if it was organized as a single distirict and operated all grades K-12. There are some districts which tuition their high school elsewhere which I did not include in the comparison.

  5. Sean

    Keene is the only regional school district in your list, think maybe a regional system has different costs?

    • The only comparable cooperative district is Keene. Being cooperative does not affect the costs of operating our middle and high schools. The more important cost factors are those districts who have all three levels of schools that they themselves run with a similar student population.

      • Rob Collins


        Both coops close to mass border. These failed to make it into their analysis. I’m sure you can guess why…

      • More dishonest argumentation.
        What is your point? Do these districts have higher measurables than Timberlane? Are they comparable? In what way?
        Truth is, you don’t know because both these districts are cooperative but each town in these districts has separate elementary schools with their own board and separate financials. They are not comparable because they don’t have consolidated financials or governance as Timberlane does.

        Furthermore Winnacunnet is only about 2000 students adding all the elementary schools together.

      • Rob Collins

        The infrastructure of these districts is much closer to what TRSD has than Hudson or Salem…

        Our elementary schools have their own budgets like these, we simply vote on them differently. There’s no more overhead cost because of their setup…

        On the flip side, they do not have the ability to redistrict their elementary schools, much like TRSD. That adds cost and that much more closely reflects the staffing constraints we see here at TRSD.

      • Rob Collins

        Why didn’t you guys consider the writing and science NECAP scores in your analysis???

        Hudson doesn’t beat us there!

      • Sean

        What about transportation costs in regional versus city?

      • Sean

        what about the ability to adjust school enrollments within city elementary schools which you can”t do in regional

      • Transportation costs are specifically omitted from the department of education’s calculation of cost per pupil. We use the DOE’s cost per pupil in the comparisions unless we indicate it was estimated, which only one figure is…. just so you know.

  6. Arthur Green

    I selected NECAP reading and math because these represent foundational skills and knowledge necessary for any other subject areas. The grade 11 2014 results are available uniformly for all districts. Participation is maximum.
    NECAP science results are a year older,generally seem to represent a lower participation rate.
    In any case, I spot checked a few of the comparable districts. True, Hudson scored lower on NECAP science than Timberlane, but Salem scored higher, and Bedford scored much higher.
    There are 40 pages of NECAP results in the Timberlane report card. I picked the metrics that I consider most informative, and I’ve given my reasons for thinking so. An alternate method would be to scour the data to find isolated facts that favor my case. This method is fully on display in Mr. Collins’ comment above.

    • Arthur Green

      A word about Exeter. Exeter is in the enrollment range of my analysis, but the NH DOE does not report consolidated Cost Per Pupil and Student Teacher Ratio for the combination of Exeter Regional (which operates the High School and Middle School) and the individual towns which each have a single-school district at the elementary level. If the consolidated CPP and STR were available, I would include the SAU in my comparables.

      This assumes that the organizational, cost and staffing issues are negligible, which I do not accept simply on Mr. Collins’ assertion.

  7. Rob Collins

    So, what about Winnacunnett?

    What’s their CPP and STR?

  8. Arthur Green comments (from Donna’s desktop):

    For the benefit of people following this thread:
    Winnacunnet SAU 21 has the following school districts
    – Winnacunnet Cooperative – 1,135 students in grades 9-12
    – Hampton Falls district, 264 students in grades K-8
    – North Hampton district, 433 students in grades K-8
    – Seabrook district, 730 students in grades K-8
    – South Hampton, 63 (yes, 63) students in grades K-8
    Total enrollment: 2,625

    I have stated several times that my selection of comparable districts is in the range of 3,000 to 5,000 enrollment. Winnacunnet is too small to be a valid comparison even if it were organized as a single district.

    NH DOE reports critical metrics by school district, not SAU.

    I could spend days getting the financial and reports for these districts, pulling them apart and constructing an consolicated equivalent to the DOE Cost Per Pupil and Student Teacher Ratio. If the numbers resulted in a comparison unfavorable to Timberlane, how much credence would Mr. Collins give those numbers?

    All of my comparison data with other districts is sourced from publicly available information from NH DOE (not all posted on the web site, some information by request, such as DOE-25 filings).

  9. Rob Collins

    OK. I did the analysis and it didn’t take days. It’s simple on a spreadsheet, you just add the single town school districts into the larger cooperative.

    Winnacunnet has a total enrollment of 3,591 (you forgot Hampton). Fits into your model nicely. I also added Exeter Cooperative with enrollment of 5,134.

    If you include the Writing and Science NECAP you can now exclude Hudson, Keene and Merrimack from your “leading cohort,” albeit as flawed as it is… I’m excluding them because you did make a point to say you wouldn’t want to hurt the district and model after other districts who performed less favorably than TRSD on every exam.

    The new “leading cohort” is Bedford, Exeter and Winnacunnett. The only district of your “leading cohort” scoring better than TRSD on all four of the NECAP exams turned out to be Bedford. Wrapping in Exeter and Winnacunnet, two cooperative districts who are much better representations of what TRSD actually is and did score better than TRSD on all four exams, the results are as follows.

    These are 2012-2013 numbers as that is the last year we have a full data set and you wouldn’t want to mix them I’m sure of, correct?

    Average STR of the new “leading cohort” = 12.8
    TRSD is 12.1

    Average CPP of the new “leading cohort” = $13,409
    TRSD is $13,329

    We are a little low on STR but right in line with the leading cohort for CPP.

    • Arthur Green posts using Donna’s userID:

      So much misdirection in Mr. Collins’ comment. But let me start by acknowledging my error. I did leave out Hampton School District in my posting which listed the schools making up Winnacunnet enrollment. The cause of this error is instructive for the discussion. Winnacunnet is a cooperative district managing one high school. It is in SAU 19. The NH DOE web site for SAU 19 lists the associated school districts which manage the elementary/middle grades, and these are the districts I mentioned in my post. Hampton is SAU 90, with a separate superintendent, managing 3 elementary schools.
      So I apologize for my error in not digging to deeper on Winnicunnet to explicitly confirm the participants in the cooperative.
      Going back to my original criteria, however, I stated explicitly that I selected all the districts which run K-12 under a single district organization. Winnacunnet does not. Exeter does not. Mr. Collins’ comparison assumes that the organizational and cost differences between Timberlane versus these multi-district and multi-SAU cooperatives are not material. I do not know how he can claim this. Separate boards, separate SAUs, and we are comparing costs to Timberlane?
      Mr. Collins states that combining the Cost Per Pupil (CPP) and Student Teacher Ratio (STR) is simple, just “add the single town districts into the larger cooperative”. He did not provide his CPP and STR numbers for these cooperatives, much less the details of his calculation. Has he had this calculation validated by the NH DOE? Perhaps not, since it is not posted on the “support” facebook, which prides itself on posting only officially validated information.
      Mr. Collins misrepresents me with the statement “you did make a point to say you wouldn’t want to hurt the district and model after other districts who performed less favorably than TRSD on every exam”. What I have said is that NECAP Grade 11 reading and math represent a broad-based and objective measure of foundational skills and knowledge. I expect a group of similar districts to have a variety of strong and weak points.
      We are trying to establish an overall level of resourcing which will enable a district to deliver a quality education. The five “Leading districts” I have identified are clearly delivering a quality education, and doing so at a significantly lower cost per pupil.
      Mr. Collins is welcome to build his own model of cost comparisons based on finding someone who is spending more than Timberlane.
      I repeat that ALL my comparisons with other districts are based on publicly sourced information. Where I use estimates, I do so to assess where Timberlane is and where it will be in advance of official numbers which may be 2 years old. Bringing me to the next instance of misdirection.
      Mr. Collins resurrects 2012/13 numbers for CPP and STR on the argument “that is the last year we have a full data set”.
      Of course, we do have a full set of CPP and STR comparisons for 2013/14 from the NH DOE, and we’ve shown that Timberlane costs jumped in just the one year. We are discussing the costs associated with the 2015/16 budget, and the past Chairman of the School Board is arguing that our costs were not out of line in 2012/13. This shows that Mr. Collins is more interested in disguising and distracting from the runaway cost increases rather than forthrightly explaining to the residents why he thinks his spending program merits support.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s