Study: N.C. charters get better results for less money

Students in N.C. charter schools earned higher reading and math scores in 2011 than their counterparts in traditional public schools,  while the charter schools got less money for doing it,  according to a new study from the University of Arkansas Department of Education Reform.

The latest study,  "The Productivity of Public Charter Schools,"  piggybacks on an April report that compared per-pupil spending on charters and other public schools.  It compares scores on the 2011 National Assessment of Educational Progress eighth-grade reading and math exams for the two groups and links those to spending.

The report,  which looks at all states that had charter schools in 2011,  shows that N.C. charter school students averaged 13 points higher in reading and nine points higher in math than students in N.C. school districts.  Meanwhile,  charter schools averaged $8,277 per charter student compared with $9,999 per district student.  The study does a lot of other number-crunching but that's the gist:  Higher scores for less money.

Skeptics may assume that's because charter schools are working with the students who tend to score higher.  But according to this study,  the N.C. charter schools averaged slightly higher percentages of low-income and disabled students than public schools across the state.

Of course,  there are plenty of caveats to consider,  and the 43-page report explores many of them.  This is one year's performance  (a year that precedes North Carolina's charter school expansion)  for one grade level.  As the study notes,  those students may have experienced a mix of charter and traditional public schooling  (and,  for that matter,  private and home-schooling),  all of which contributes to eighth-grade scores. The report uses that data to extrapolate a  "return on investment"  based on lifetime earnings.  I'm skeptical of that technique,  which is used to turn small data points into huge savings by any number of educational groups,  including traditional public schools.

The researchers note that the overall analysis leads to one clear national finding:  "Charter schools tend to exhibit more productivity than traditional public schools."

You can bet that will come up as North Carolina debates how to balance its investment in various forms of public education.

Baker

Update: A reader steered me to a University of Colorado National Education Policy Center review of the April report on charter inequities. Reviewer Bruce Baker of Rutgers University says the University of Arkansas study  "displays complete lack of understanding of intergovernmental fiscal relationships."  For instance,  he writes,  money that is passed through school districts for distribution to charters is counted as school district revenue in per-pupil calculations  (CMS passed through about $23 million in 2013).

"In addition, the report suffers from alarmingly vague documentation regarding data sources and methodologies, and it constructs entirely inappropriate comparisons of student population characteristics,"  Baker writes.  "Simply put, the findings and conclusions of the study
are not valid or useful."

As some of you have noted,  and as I pointed out in the post about the April report,  the University of Arkansas research is part of the university's School Choice Demonstration Project,  which is funded by the Walton Family Foundation.

Read more here: http://obsyourschools.blogspot.com/search?q=+university+arkansas+charter#storylink=cpy

Read more here: http://obsyourschools.blogspot.com/2014/05/report-nc-charter-schools-dont-get-fair.html#storylink=cpy

Report: NC charter schools don't get fair share

North Carolina's charter schools averaged $1,722 per pupil less than the state's traditional public schools in 2011,  according to a new report from the University of Arkansas Department of Education Reform.

"Charter School Funding:  Inequity Expands" follows up on similar national studies done by other groups, comparing funding in 2003 and 2007.  The latest study is part of the university's School Choice Demonstration Project and is funded by the Walton Family Foundation.

"Since public charter schools are becoming increasingly politically popular and therefore common in the U.S., we might expect that they would be funded at levels comparable to traditional public schools. After all, they are public schools, too,"  the authors write.  "We would be mistaken."

The researchers found a national gap of 28.4 percent,  with charters averaging $3,814 less per pupil from combined state,  local,  federal and private money.  In North Carolina the gap was 17.2 percent,  or $8,277 per charter student compared with $9,999 per district student.

As charter advocate Larry Bumgarner and I have been discussing in recent comments,  per-pupil funding is a complex and confusing system.  The university's report on North Carolina provides a detailed explanation of state funding formulas. "In practice charter school State revenues are nearing parity with district per pupil State revenues, which is significant,"  the report notes.

The N.C. breakdown cites causes of the funding gap, including the absence of local money for charter school buildings and a lower level of federal funding for charters.  (Note that this study landed during the recession,  when federal stimulus money was coming in to local districts.  It also came before North Carolina's charter school expansion,  which might or might not affect per-pupil spending.)

The study does a breakout on Wake County charters and district schools,  but there's nothing on Mecklenburg,  which has the largest charter school enrollment.  I was interested to see that statewide,  charter schools served a slightly higher percentage of low-income and special-education students than district schools.  As the report notes,  that can vary widely by school.

The report gives North and South Carolina D grades for charter/district discrepancies.  Tennessee,  where charter schools averaged $15 per pupil more than district schools,  got the only A.

Also on charters and district schools:  I'm moving toward getting payroll data from area charters,  Charlotte-Mecklenburg Schools and nearby districts.  Expect more on that after the Observer/PNC forum on teacher compensation,  which takes place Monday evening.  Seats are still available;  click here for details and to reserve a seat.

County education money: Whose is it?

A sign of the changing education landscape:  For the first time in 13 budget cycles I've covered,  Mecklenburg County commissioners sat down to talk about charter schools.

Charlotte-Mecklenburg Schools,  which briefed the commissioners,  has posted that report online. It includes a lot of interesting items for anyone who cares about the issue,  including a map of where Mecklenburg's charter students live,  a list of which schools they attend and a demographic comparison between CMS and the students attending Mecklenburg's charter schools.

The presentation got me thinking about the way I've always reported on county funding as an allocation for CMS.  There's good reason for that.  The superintendent,  top administrators and school board spend weeks crafting a request,  which commissioners scrutinize and vote on.  While everyone knows that there's a pass-through to charters,  it's CMS that gets the scrutiny, praise and heat.

Given the work that CMS puts into getting that allocation,  it's easy to understand why officials and supporters might resent about being forced to give up some of their money  --  especially when that pass-along is expected to top $30 million next year.

But if you look at is as the county's allocation for public education,  it makes more sense.  If CMS projections materialize,  about 157,000 Mecklenburg students will enroll in public schools next year.  CMS will get about 91 percent of them, and 91 percent of the county money.  The charters that serve the other 9 percent will divvy up that portion.

The big question is whether that's a smart way to spend local money. CMS board member Eric Davis took issue with my recent post critiquing his comparison of CMS and charter spending projections.  I said it was a false premise to assume that the per-pupil share passed along to charters meant anything about the actual cost of operating those charters.  Davis said I missed his point:

Davis

While there are many charter schools that are well performing, my comments were solely based on financial efficiency of charter schools versus CMS.

CMS is better positioned to absorb another 2300 students within our existing network of 160 schools than opening 10 charter schools, each of which will need a new principal and other support staff. That staff already exists in CMS.

Moreover, CMS per pupil funding continues to go down, dropping over the past 5 years by 7.2% in state funding and 5.4% in county funding. Since charter schools receive the same per pupil funding as CMS, they offer no possibility of reducing per pupil expenditure when another charter school opens. Hence, charter schools are a less efficient way to educate an expanding student population than the existing system.

In a time when we hear that our state does not have money for public education due to other issues such as increased health costs, it would seem that our decisions would be guided by how to improve the efficiency of the existing system. Charter schools, while they provide choice, do so in a more expensive way than increasing choice within the existing system.

Davis might be right  --  if CMS were,  in fact,  planning to absorb all the new students into its existing schools.  But CMS is doing exactly what Davis and Superintendent Heath Morrison have criticized charters for doing:  Creating small schools.  They're counting on the state to approve $922,000 in the 2014-15 budget to launch the UNCC Early College High  (100 students),  Levine Middle College High and Harper Middle College High (69 students who have so far applied for next year's debut on a combined campus).  That's in addition to the local and state money CMS would normally spend for those students,  and the county money to set up mobile classrooms.

CMS is also creating Hawthorne Health Sciences, a magnet high school with 87 students enrolled so far,  and a new Montessori magnet in Huntersville with 92 enrolled so far.

The state requires a minimum of 65 students in each charter school.  Even the small charters generally try to open with more than that.  And most plan to add grades and build enrollment in coming years,  exactly as CMS plans to do with its small start-ups.

Bottom line:  Large schools generally have lower per-pupil costs than small ones,  whether they're charters or CMS.  Schools with low poverty and few kids with special needs are cheaper than those at the other end of the spectrum,  regardless of who's running them.

And CMS is opening small schools for exactly the same reason the state is authorizing charters:  They believe it's the right thing for students.

When I pressed Morrison about the per-pupil cost of his small schools,  he noted the success of existing ones such as Cato Middle College High.  "How do you put a price tag on the overall quality?"  he asked.

That's a question that bears thoughtful exploration,  for charter and district schools alike. 

CMS raises and charter costs: A math challenge

School board members delved into math questions about employee raises and charter school costs at Charlotte-Mecklenburg Schools' first budget work session Tuesday.

Tuesday's presentation  (CMS plans to post it on the budget page today)  included a preliminary suggestion of $17 million to $18 million in county money to provide raises for all employees.  A board member asked what size raise that would cover.

Image: Cambridgecoaching.com

Superintendent Heath Morrison and Chief Financial Office Sheila Shirley said it's too early to say.  Most CMS employees are paid by the state, but others are paid with county,  federal or grant money.  When the state gives an across-the-board raise,  CMS traditionally uses county money to make sure all employees get that raise.  So if state lawmakers were to approve a 2 percent raise,  CMS would need $4 million to $5 million in county money to match it,  they said.

On the other hand,  if the state provides nothing,  CMS would spend about $18 million to give employees a 2 percent raise with county money alone,  according to the presentation.  Plus,  Morrison noted that the plan to boost starting teachers' pay proposed by Gov. Pat McCrory and legislative leaders may bring yet-unknown county costs.

In that case, a vague answer was the right one.  But when it came to the cost of sending students to charter schools vs. CMS, a teacher should have stepped in.

The presentation projected 2,296 additional charter students next year,  with CMS passing along $7.5 million in county money.  It also said CMS would need $400,000 from the county for an additional 754 CMS students.

New board member Paul Bailey asked whether such disproportionate numbers could possibly be right.  It was a good question.  The charter allotment breaks down to about $3,267 per student, while the CMS total is $530.50 per pupil.

Morrison and Shirley said the costs aren't directly comparable. They explained the system that requires CMS to pass along a per-pupil share of the county allotment for each Mecklenburg student who attends one of the independently-run public schools.

Davis

When Eric Davis got his turn,  he looped back to those numbers,  saying how much more cost-effective it is to send students to CMS than to create new schools for them.  Describing himself as a former Republican,  the unaffiliated Davis said he would be aghast at spending so much to finance new charters when the state is  "starving"  traditional public schools.

It was a dramatic speech based on a false premise.  The per-pupil pass-along to charters has nothing to do with the cost of operating those schools,  as Davis most likely knows.  In fact,  that cost is precisely and by definition the same as the average cost for CMS students, new and existing.  That's how it works:  However big the county  "pie"  for education is,  each student gets a proportionate slice.  The only thing that drives up the size of a charter student's  slice is county commissioners' generosity in responding to the CMS request.

I'm not sure exactly what's included in the $400,000 estimate for the addition of 754 CMS students,  but it's clearly not the full county cost for their education.

Many of us would love to know which schools deliver the best academic value for the dollar.  But looking at those two lines in the CMS presentation won't give us the answer.

Fair share from CMS?

Mecklenburg County Commissioner Matthew Ridenhour last week asked Superintendent Heath Morrison for a breakdown of Charlotte-Mecklenburg Schools spending by voting district. Such a breakdown,  he said,  might help address concerns that lead some people to talk about splitting the countywide district into smaller ones.

Ridenhour

Ridenhour, a Republican elected last year,  is jumping into a longstanding debate over the fairness of how and where CMS spends its money.  The gist:  Schools in low-income neighborhoods get extra money to help overcome student disadvantages.  They tend to have higher per-pupil spending than suburban schools,  partly because of that extra aid and partly because suburban schools are generally larger,  spreading basic operating costs among more kids.  But residents of more affluent zones tend to pay more in property taxes,  leading to rumblings about unfairness.

Morrison,  who started the CMS job in July,  said he'll try to answer Ridenhour's question.  I'll be curious to see what he comes up with.

His predecessor,  Peter Gorman,  calculated per-pupil spending at each school as part of a CMS equity report.  The county has used those numbers to create a per-pupil average for each district.

But Morrison's crew didn't do an equity report and hasn't released updated per-pupil spending numbers.

Dunlap

And as County Commissioner George Dunlap,  a Democrat and former school board member,  noted,  school locations don't correspond to services that residents of a voting district receive.  Many students live in one district and attend school in another,  especially if they're in magnet or alternative schools.

"This is one community,"  Dunlap said.  "I don't think we ought to be trying to split it up by district."

One of the drawbacks of covering education for more than a decade is that some of the back-and-forth starts to feel like watching an old married couple argue.  County Commissioner Bill James, an accountant and a Republican, has been arguing for years that CMS gets too much money and doesn't provide enough results.  This time around,  he didn't seem to find the energy for critiquing the numbers.

James

"I just don't really feel that educational achievement is getting better,"  James said after watching a presentation on CMS academic gains.  "Maybe it's a lack of PR on the part of CMS."

"Feelings are not facts,"  responded Dunlap.  "Just because you feel a certain way doesn't make it true."

Dunlap urged his colleagues to look at the data and see how much progress CMS has made toward narrowing the performance gaps between black, Hispanic and white students and between poor and middle-class students.

At the risk of being a party-pooper  --  and the even bigger risk of getting in the middle of a political spitting match  --  I'd note that those numbers aren't as meaningful as they look.  That's because the CMS charts compare results from 2008,  when students took state exams only once,  with those from 2012,  when students who failed the first time retook the test.  The state launched that requirement in 2009,  and the result was an immediate jump in pass rates.  Groups that had more students falling just below the grade-level cutoff  (such as black, Hispanic and low-income students)  saw big gains,  while the change was smaller for groups where most students passed on the first try  (white, Asian and middle-class students).

At the time, Gorman blasted the retesting as artificial inflation of results.  For the first couple of years he offered comparisons of pass rates before and after the retest bump.

That's probably not practical now.  But if CMS wants to make a fair comparison,  all it has to do is use 2009,  rather than 2008,  as the baseline.  If the gaps have still narrowed,  it says something about student achievement,  not just changing rules.

CMS defines key school budget terms

With CMS studying an array of cutbacks to plug a budget gap of as much as $100 million, many parents and civic leaders are paying close attention to the unfolding budget deliberations. The school board is slated to talk budget issues at its meeting tomorrow evening, weather permitting. CMS' recent release of per-pupil, per-school expenditures generated a lot of interest last week. The list included expenditures for "school activity fund," a term that has left a lot of readers scratching their heads. What, they asked, does that exactly include? Superintendent Peter Gorman late last week e-mailed school board members a document that explains that term, along with weighted student staffing and other budget items that figure to loom large during this spring's budget discussions.

Perhaps more importantly, the document Gorman gave the school board also provides an overview of the various factors that can influence per-pupil spending. Considering the intensity of the debates that always erupt surrounding questions of equity (and the fact that some of the poorest schools receive twice as much as some of the wealthier schools), those factors are worth studying. You can take a look here.

Eric Frazier

Geek alert: CMS per-pupil spending

CMS has posted its latest chart of calculations on per-pupil spending by school, as board members prepare to start figuring out where to slash tens of millions of dollars for the coming year.

Eric Frazier, efrazier@charlotteobserver.com, is doing the story on the numbers. Since we're still working getting him hooked up on this blog, I'll share the good and bad news for those who like to do their own number-crunching.

The good: CMS is sharing great info about spending, academic performance, student/teacher ratios, teacher experience, poverty and race. And they've included enough detail that you can see how the staff did their calculations.

The bad: It's nearly impossible to read on the PDF files they've posted. (If you know any tips for making that work better, please share.)

The good: The public information office sent us an Excel version and is checking out the N.C. Department of Public Instruction's system that lets users download data in Excel.

Eric and I can pass along the Excel version with those of you who want to study the chart. Contact Eric if you want to discuss issues for the story; if you just want the geek sheet, I can help at ahelms@charlotteobserver.com