I will protect your pensions. Nothing about your pension is going to change when I am governor. - Chris Christie, "An Open Letter to the Teachers of NJ" October, 2009

Friday, June 30, 2017

The Charter School Free-Riding Problem

There's a piece by NY Daily News Editorial Board member Alyssa Katz that's been passing around the edu-bloggosphere to the great consternation of charter school cheerleaders. The piece points out, in personal terms, an issue that those of us who study charters have known about for a while:
The usual end-of-year school rites for our fifth-grader are especially bittersweet this year, because it is her last at the charter school where she has learned, played, made friends and grown since kindergarten.
Unlike many kids at her stage, she had a choice to stay at her K-8 school — but as a family we together decided to jump from the charter to a district junior high run by the city’s Department of Education.
Some extracurricular forces eased the choice. My husband, who’s logged hundreds of miles driving to and fro, will hand our girl off to a convenient bus. She in turn will be thrilled to shed a loathed uniform. Me, I look forward to an end to lunch box prep, thanks to an improved cafeteria menu.
But the bottom line is that her elementary-school years were marked with a whirlwind of teachers that, if she and her classmates were lucky, would last the year and then move on.  
The ritual became as certain as winter succeeded fall: Some parent would post on the school Facebook group that their child’s teacher was leaving mid-year. Moans and commiseration ensued.
Our child avoided that fate until last fall, when, two weeks in, her promising teacher — a veteran at three years served — suddenly vanished, and a substitute arrived much sooner than any explanation. Her class rebound its footing, eventually, with a new teacher — but never quite recovered from those lost weeks. [emphasis mine]
What Katz describes here is quite typical for charters around the nation. In my report on New Jersey charter schools' finances and staffing, I found clear distinctions between charter and public district educators: charter teachers have less experience, are less likely to have an advanced degree, and are paid less, even when accounting for experience.

There are exceptions. As I've noted before, the largest NJ charter networks, such as TEAM-KIPP and Uncommon (North Star), pay teachers more at the outset of their careers. But the lack of experienced teachers on the staffs of these charter networks suggests they don't hang around long enough to make the high-five figure salaries* district teachers start to make around their second decade in the profession.

The response I've had to this from charter cheerleaders, such as Peter Cunningham -- here writing in the News in reply to Katz -- is usually: "So what?"
Daily News editorial writer Alyssa Katz has decided to pull her daughter out of a public charter school in New York City because of high teacher turnover. Katz cites a statistic that teacher turnover in the city's charters is more than twice as high as in unionized district schools, which generally pay more, have shorter hours and, of course, provide tenure protections and other benefits.
To her credit, Katz acknowledges that New York's largest charter school organization, Success Academy, has "skyrocketing" test scores despite high rates of teacher turnover, which raises the question of whether teacher turnover is good or bad.
The more important question is, however, should we even pay attention to teacher turnover?
In theory, unions produce happier, more secure teachers and they, in turn, produce better educational outcomes. In practice, that's not the case. Lots of unionized teachers are very unhappy, enormous numbers of unionized teachers leave the profession, and lots of unionized schools get awful results.
I'll interrupt here to note that if unionized teachers are unhappy and leave the profession, there's little evidence it's because of unions themselves; "reform" and a lack of funding appear to play a much larger role in teacher dissatisfaction. Also, there is no evidence I am aware of that shows a negative correlation between school performance and unionization; research suggests quite the opposite, in fact.

Cunningham continues:
Too often, in the ongoing education debates, charter opponents will talk about anything except results. They will harp on "no excuses" discipline practices. They will attack charter funders. They will complain about public charters draining funds from public schools.
They won't talk about test scores, unless, of course, the scores are bad, in which case they will cite them and call for closing down low-performing charters. [emphasis mine]
Fine, let's talk about results:

In the aggregate and across the nation, there is little evidence charter schools get, on average, significantly better results than public district schools

For example, the CREDO studies, long cited by charter advocates, show on average that the gains of the charter sector are, at best, quite small (even if CREDO oversells them by using a conversion to "days of learning" that is wholly invalidated).

There are pockets, however, such as Boston, where charter gains are larger. To be clear: the gains almost never come close to "closing the achievement gap," no matter how some in the credulous press choose to spin the results. But there are gains. The problem is that too many reformy folks want to end the conversation right there; what they should be asking is:

1) Why?
2) At what cost?

Over the past several years, it's become clear to me what the answers to #1 are:

A longer school day and school year, smaller class sizes, and one-on-one student tutoring.

This is the resource-intensive part of "no excuses" chartering that seems to always get lost in the conversation. Well, more accurately: it's never fully explored. Because everyone loves the idea of more teachers working a longer day and year for at-risk kids -- but they don't really like talking about the cost.

Basic economic theory suggests that getting people of equal qualifications and effectiveness to work a longer day and year will cost an employer more in wages. So charter schools have two choices for getting more teachers to work longer: they can pay more, or they can recruit less expensive staffs.

We've already noted that some large charter networks (and even some smaller ones) rely on their ability to gather philanthropic contributions to supplement the public funds collected so they can increase staff hours/days, and reduce student-to-teacher ratios. But the other way they can gain a resource advantage is to maintain a less expensive staff by constantly churning it. There is, however, a problem...

Teachers, like so many other professions, gain increases in wages through accumulated experience. A teacher who has 20 years of experience in the same district will almost always make more than one with only 10. Some have made the case experience and salary should be decoupled, but there's more chance of my winning the next Olympic silver medal in the backstroke than of that happening. We'd have to either cut experienced teacher wages and distribute them to inexperienced teachers, or pour a lot more money into the system; neither is going to happen in this world.

When teachers enter the profession, they are well aware of how the step guides work. Which means they are already acclimated to the idea that they will make considerably less in their earlier years with the reward of better pay later on.

As a teacher, I can also add that my personal experience with younger teachers is that they expect their first few years are going to be much rougher -- in terms of their time commitments, their stress levels, their choice of assignment, etc. -- than what will follow later in their careers.

Which brings us back to the critical section of Katz's piece:
The big reason for charters’ turnover plague is plain as day: District school teachers are universally represented by teachers unions, and enjoy contracts whose ample benefits include generous pension plans, non-negotiable business hours and tenure. 
When our child’s teacher got an offer on Long Island last September, that was that. 
Charter school teachers, in glaring contrast are often called on to work extra hours after school, and during summers, and whenever. 
Which job would you pick if given a choice? Not even a close call. For all but the Teach for America types who intend to log a few years and switch tracks, the union jobs are better jobs, where educators build careers.
The data we have on charter teacher attrition is pretty thin, but we do know a few things. The teacher attrition rate at high-profile carter networks like Success Academy is very high compared to the NYC public schools, and switching between schools within the network doesn't account for the disparity. Charter teachers do often cite working conditions and job dissatisfaction as reasons for leaving their schools. Chris Torres finds that when a teacher perceives that their workload is too great at a charter, that teacher is more likely to leave (although there appears to be an interaction with teachers' perceptions of their schools' leadership).

We need better empirical evidence to back up a claim that charter schools are serving as springboards for teachers in search of unionized jobs in public district schools. But the anecdotal evidence is certainly piling up:
One veteran charter school teacher who requested anonymity, wary of her school’s response to her comments, says that even as schools use teachers, teachers use schools—as way stations to other, long-term goals. Often, she said, new teachers work for a year or two at a charter school to buff up a resume, ahead of a search for a union job. “People go there [to charters] until they can get another job. It’s a stepping stone to a teaching career, to a union job with benefits, like vacation, and tenure.” [emphasis mine]
There is, in my opinion, at least enough evidence so far to suggest this is a plausible theory. Which brings us to the problem of "free riding." Martin Carnoy explains:
The “free rider” aspect of teacher costs in private schools, whether voucher or charter, means that the supply of young people entering the teaching profession is maintained by the salary structure and tenure system in public education. Without this structure, many fewer individuals would take the training needed to become certified to enter teaching. Since teaching salaries are low compared with other professions, the prospect of tenure and a decent pension provides the option of security as compensation for low pay. This pool of young, trained teachers is available to voucher and charter schools, generally at even lower pay than in the public sector and without promise of tenure or a pension, but with the possibility of training and experience. Thus, the public education employment and salary system “subsidizes” lower teacher costs in private and charter schools. In other words, for private schools to have lower costs, it is necessary to maintain a largely public system that pays teachers reasonable (but still low) salaries and provides for a teacher promotion ladder and job security. [emphasis mine]
It's telling to me that Katz's teacher took a job on Long Island. The ultimate goal wasn't just a public school job; it was a job at a particular type of public school.** We know that teachers in urban schools, in schools with greater proportions of economically disadvantaged students, and in schools with worse academic performance are more likely to move to a new school.

What is likely happening now is that the pattern of teachers leaving urban schools for what they perceive as less stressful jobs is being amplified by the larger work demands and smaller pay of charter schools. 

Urban public school teachers might have been more likely to move to another district before, but at least their school days and years were similar to more affluent districts. Charter teachers, however, don't even have that in common: they get even lower pay and even longer hours.

Urban teachers might also have stayed in their districts because they earned tenure, which they'd have to give up when changing districts. Charter teachers often never gain tenure. So why would they ever stay if a better paying, less stressful, more secure gig comes along?

What Cunningham, with his focus on "results," fails to see is that the problem here is bigger than we'd think by simply looking at some -- some -- charters' somewhat better test scores. First, there's good reason to believe a model like this can't be sustained as charter proliferation grows.

Second: The charter school model depends on public schools subsidizing the free riding of charters. If that free riding isn't accounted for in resource distribution, the charters are creating patterns of inequity.

Charter cheerleaders often talk about what "high-performing" charters can teach public district schools. I'd argue that if "high-performing" charters are teaching us anything that can be applied to the entire system, it's that resources smatter. Charters may get those extra resources through things like philanthropy -- or they may artificially gain a resource advantage through churning an inexperienced staff that is ultimately subsidized through the public, district schools.

In either case, it's an unfair advantage, and it needs to be acknowledged before anyone makes the case that charters "do more with less." Increasingly, the evidence suggests the few charters that "do more" are doing so because they have more.

* Re-reading that sentence heightens its absurdity: "Ooh, high-five figures! In New Jersey! For college-educated professionals! After 15-20 years! What a deal!"

The teacher pay penalty is real and it cuts across the entire profession, not just the charters. But as poorly paid as teachers may be, charter teachers are paid even worse.

** OK, school districts on Long Island do vary considerably in their resources and their student populations. We don't know exactly where Katz's child's teacher wound up. 

Monday, June 12, 2017

Should Families Really Have to "Choose" Healthy Schools?

Looks like it's going to be a big week for school "choice." The National Charter School Convention is in full swing (I've been following the hashtag #ncsc17, and a better "reform" cliche lexicon you will not find). The folks at CREDO released yet another charter school report that once again invalidly translates small effect sizes into "days of learning" in a vain effort to show charter schools are full of chartery awesomeness.

And the doyen of "choice," Eva Moskowitz, has apparently picked up the coveted Broad Prize for her work in expanding a network of charter schools with practices so "innovative," she wants to put them on a digital platforms and share them with the world.

If there was any justice in this world, that video would have been running in a loop behind Moskowitz as she accepted Eli Broad's dough...

As those of us who follow education policy and live in the greater-NYC area know, Moskowitz's Success Academy has benefitted enormously from philanthropic giving. $8.5 million from a hedge-funder in 2015. $9.3 million in one night later that year. $25 million from another in 2016. Plus another $10 million from some pikers...

This is all in addition to the monies SA gets from the government for the students it enrolls. Moskowitz's powerful friends have even made sure that she doesn't have to play by the rules that everyone else must follow. The result is a school system swimming in money -- a system that relies on funds that no one else gets to access.

Understand, it's not just the kids on SA's "got-to-go" list who miss out on all this Wall Street largesse; any NYC student, public or charter, who isn't in Moskowitz's network misses out on the benefits of all this extra cash.

In the case of Success Academy and other big-profile charter networks, the benefits of lots of extra resources come at the cost of not having your child enrolled in a democratically and locally controlled public school, with greater transparency and greater access to due process and student/family rights.

Why I am bringing all this up today? Well...
PATERSON – Monday's heat wave prompted city school officials to send elementary students home at 1 p.m. on Monday.
High school students already were getting out early because of exams, according to district spokeswoman Terry Corallo. The district has more than a dozen schools that are more than a century old and lack air conditioning.
Staff members on Monday were required to stay at work until after 3 p.m., prompting criticism from the president of the teachers union. [emphasis mine]
And it's not just Paterson:
Thermometers are rising and more than 20,000 students in public schools in PlainfieldTrenton, and other districts throughout the state are being sent home early over the next two days. 
With the pressure of finals in the air, many students and school employees also have to contend with rising classrooms temperatures.
Few examples so elegantly show the wide disparities in school conditions in New Jersey.
In some districts, the rising temperatures won't mean much and the learning process will continue unabated.  In other districts, schools will be forced to shutter and students will lose precious hours of instruction. 
In what is often a clear divide between affluent and poorer districts, some students and school employees will learn in comfortable climate controlled classrooms, while others will struggle to learn and teach in classrooms with temperatures approaching and sometimes exceeding triple digits. [emphasis mine]
It would be an overstatement to say that every classroom in every affluent district in New Jersey has A/C; I know of several examples personally where that's not the case. But there's no doubt a student in the leafy 'burbs is much more likely to have A/C in her school than a child in an urban public school...

Unless that child is enrolled in a well-funded, well-connected charter school. From 2015:
Meanwhile, in Camden, the aunt of a student at Bonsall Elementary School posts a video (which I can't embed here because it's on Facebook, so click the link to watch) showing how students on the two lower floors are sweltering in classrooms with no air conditioning.

But up on the third floor, it's nice and cool. Why? Because that floor was taken over by the Uncommon charter chain, which somehow allowed the district to magically acquire the funds necessary for the school's renovation. Except somehow, when it came to HVAC needs, the floors housing classrooms for the public district schools didn't get refurbished in time for the start of the school year.
Bob Braun has been reporting on the disparity in A/C between charters and public district schools in Newark for years. In addition, when the new, modern Teachers Village was constructed in Newark, its three school spaces went immediately to charters; NPS schools were left to wither in the sun.

NYC Mayor Bill de Blasio, to his great credit, has made getting air conditioning units into NYC schools a priority. That's great, but I have to ask: why the wait? Why do charters like SA have access to millions of dollars in extra funding to keep their kids cool while over one-quarter of NYC classrooms swelter?

Why do Paterson public school students have to end their school day early while New Jersey gives out millions in financing to a publicly-funded charter that pays rent to a private entity? Why was there plenty of money for Teachers Village, occupied solely by charter schools, but scant few dollars to renovate Newark Public Schools' aging infrastructure -- including a plumbing system full of lead? Why did charter students in Camden get air conditioning while CPS students in the very same building did not?

No family should have to "choose" a healthy school for their child. If you really believe, as many of the fine folks partying it up at the NCSC Convention tonight apparently do, that we need more "cooperation" between district and charter schools, why would you stand for a school funding system that advantages high-profile, well-connected charters over public district schools and mom-and-pop charters that may not have hedge fund-types writing them big checks?

Again, there is no question that more affluent public school districts across the country have been unfairly enjoying a resource advantage over less affluent districts. But allowing vastly wealthy people to pick and choose which charter school networks they like, and then setting those schools up with both "no excuses" discipline and A/C, hardly seems like an equitable plan.

Rather than picking a few urban charter schools Hunger Games-style to get decent facilities, why don't we instead tax the donors to Success Academy a few percentage points more and use the money to make sure all schools are safe, clean, and healthy? 

Is anyone really against that?

h/t UFT

ADDING: Via Twitter, NPS staff report 100+ degrees today in some schools, yet no early dismissal.

Golly, I wonder what the temperature in Chris Cerf's office was...

I have always stated that the Paterson State Appointed District Superintendent Dr. Donnie Evans never fully understood the adverse impact that these inhumane conditions have on our students and our employees. I believe that this is especially true as he works from his air-conditioned 4th-floor corner office located at 90 Delaware Avenue. Be this as it may, while sorting through the OPRA request, I could not help but notice one particular document and the message it sends to our students and employees.
According to district records, on July 26, 2016, a receipt was paid in the amount of $250.77 for the following service/repairs, “AC not working, needs service.” According to these same records, the air condition repairs were made to a vehicle listing Dr. Donnie Evans as the driver. For those who do not know, Dr. Evans is provided the use of a District school vehicle. The irony here is that the repairs listed on the invoice I am referencing is for the same luxuries Evans has denied the students and staff without air conditioning for years.
A reminder: Evans serves at the pleasure of Governor Chris Christie, who wants teachers to work in the summer for no extra pay in classrooms with no A/C.

But not his state-appointed superintendents. Lovely.

Saturday, June 10, 2017

New Jersey's Totally Screwed Up Charter School System, Clifton Edition

Here we go again:
CLIFTON — City school board members remain united in opposition to the state education department's mandate that the district allocate $2 million toward a controversial charter school group.
The state Department of Education issued a directive to the Clifton school board late in the budgetary process, said Clifton school officials.
The directive calls for the funding of 225 students set to attend Passaic Arts and Science Charter schools, including a new building opening this fall on Clifton Avenue. The school formerly housed the Sacred Heart Elementary School.
The PASC is affiliated with the ILearn LLC Network, which is the subject of investigations due to allegations of fraud.
Business administrator Edward Appleton said the school system learned of the major increase the final week in February. The state’s directive created a spike in the board’s imposed charter school tuition from $2.69 million in the 2016-2017 budget to $4.68 million for next year, he said. [emphasis mine]
And so it continues with New Jersey's insane charter authorizing system: Trenton mandates a school district must give up funds to support a charter school that the district had no say in approving. Worse, the district cannot exercise any oversight authority over the charter: iLearn can spend the funding the state mandates the district provide any way they wish, so long as NJDOE approves.

It's worth pointing out that, according to data from the Education Law Center, Clifton has suffered from persistent underfunding of state education aid under the Christie administration: cumulatively $73.8 million since 2010. And yet the same administration mandates that Clifton taxpayers put up millions of dollars every year for a charter school the community may not even want.

The standard answer from reformsters to this complaint is that "families can vote with their feet": if the charter can't attract students, it will close. This argument fails on several levels, not the least of which is that schooling is a community concern -- not simply the concern of parents -- and that even taxpayers who don't have children in the school system have every right to expect that their hard earned dollars are being used efficiently and effectively.

And there's little reason to believe NJDOE is currently up to that task, no matter what the charters may say:
Dawn Fantasia, iLearn’s communications director, said the new school campus is located at 43 Clifton Ave., and denied Daley’s allegations that school has not publicized its financial reports.
“Passaic Arts and Science Charter School is a public charter school,” Fantasia said. “As per the NJDOE regulations and guidelines, all audit reports are made public and may be easily found and accessed with a simple search.”
She stated the charter's latest public audit, available on the state's education website, contradicts the “alleged lack of transparency purported by Commissioner Daley.”
Um, no. Calling this short document an "audit" is a gross misrepresentation. The plain fact, as Julia Sass Rubin and others have pointed out, is that New Jersey charter schools are far less transparent in their financial dealings than public school districts. There is no way to know, looking at this document or the charter's CAFR, what relationships and contracts exist between the charter, its board, and its contractors, vendors and lease-holders.

As NorthJersey.com pointed out in their extensive investigation of iLearn, not only is it difficult to get information about these relationships; the NJDOE has become a revolving door for regulators looking to make careers in charter school management:
Connections run deep among people involved with the schools, Gulenist groups and Turkish charter schools elsewhere in the U.S:
  • Two of the New Jersey schools, for example, have a founder who has served as a director at the New York-based Alliance for Shared Values, considered the voice of the Gulen movement in this country. 
  • The CEO of iLearn Schools Inc. – an Elmwood Park-based non-profit that manages four of the local charter schools – comes from a charter network in Texas that the Turkish government claims is linked to the Gulen movement.
The schools and their vendors have successfully courted prominent public-school educators and political figures.
  • The state’s top charter school regulator, Harold Lee, left his post last summer for a job at iLearn. 
  • Security consulting contracts at four of the schools worth more than $90,000 a year are held by ex-Bergen County Sheriff Leo McGuire, who took a 10-day trip to Turkey before he left office in 2010 with his family and local Turkish nationals tied to the schools. It was paid for in part by a Gulenist group.
More than $30 million in long-term, low-interest loans have been granted by the state to benefit the Paterson science and technology charter despite its continuing financial and academic troubles:

In 2014, a Wall Street ratings agency downgraded the bonds issued for its expansion to junk status because the school’s revenues had fallen. Last year, Wall Street lowered its overall outlook on the bonds to “negative.”

Tracking tax dollars spent by the schools can be difficult because of loopholes in state law:

  • ILearn, which is set to add a fifth charter to its chain this year, declined to answer routine requests for information about its payroll, saying that as a private contractor it is not subject to the state Open Public Records law. 
  • State officials said it is unclear if such charter-management organizations fall under the law, even though charters draw their funding directly from the tax-funded budgets of regular public schools. [emphasis mine]
Is anyone seriously suggesting the Clifton BOE ought to just accept all this? That they don't have a fiduciary responsibility to their constituents to make sure Passaic A&S and iLearn are using revenues appropriately -- especially when the town's public schools are being short-changed by the very state administration that forced them to fund this charter school?

Which brings us to the second way the reformster argument for laissez faire charter regulation falls apart: how is Clifton supposed to "compete" for students when it doesn't get the resources the state's own law says it needs, even as it serves a different population of students compared to Passaic A&S?

We see this over and over again: Charter schools enroll a much smaller percentage of special education students compared to their hosting public school districts. This is certainly the case here; in addition:

As is typical for New Jersey, the students with the most costly special education disabilities are concentrated in the public district schools, and not the charters. Which means that even as charters like Paterson A&S are syphoning funds away from an already underfunded district, they aren't taking the students who are most expensive to educate.

How does this manifest in spending?

The "Budgetary Costs Per Pupil" -- which the state itself says "...are the costs of governance, support, and instruction that are considered common to all school districts and generally are uniform among them" -- between CPS and Passaic A&S are about the same. But CPS spends much more on instruction and support services, while Passaic A&S puts its money into administration and plant costs. Support services help all students, but are particularly targeted to those with special education needs.

It's clear what's happening: Passaic A&S has proportionally far fewer special needs students, and therefore puts its money into other spending categories. Does anyone argue against the idea that the taxpayers of Clifton have every right to know how those monies are being spent? That they ought to know exactly where plant and administration funds flow? That the Clifton BOE has an affirmative obligation to protect its town's interests and properly regulate how charter monies are being spent? That the charter's inability to leverage economies of scale might make it a bad deal for the city's taxpayers?

Let's look at that classroom instruction category a little more, with the understanding that most of those funds are spent on teacher salaries:

I've pointed out many times that charter teachers tend to have far less experience than public district school teachers. But Passaic A&S has one of the least experienced staffs I have seen in the NJ data. Nearly 4 in 5 of its teachers have less than 3 years of experience; that's simply astonishing.

A charter school that only hires inexperienced teachers is arguably "free riding" on salaries: it takes advantage of the fact that teachers might be able to transfer to a better paying public district school later in their careers, and therefore will accept less pay now with the promise of more later. But Passaic A&S takes this a step further: it way underpays its teachers relative to CPS even controlling for experience, allowing more funds to be plowed into administration and plant costs.

Does NJDOE care about any of this? Certainly, Chris Christie doesn't; he loves Gulen-linked charter schools so much he visits them repeatedly, singing their praises while simultaneously pushing insane school funding plans that would decimate the state's urban school districts. He says these schools "do more with less" -- but he never asks how the students might differ.

I've done variations of this graph before. It takes a little explaining, but I find it very telling. What I've done here is use a linear regression model to explain variation in scores on the PARCC Algebra I exam based on several student population characteristics: free-lunch eligibility, special education percentage*, Limited English Proficiency percentage, and racial characteristics.** The model here isn't as robust as for other assessments, as only 22 percent of the variation in scores is explained. But it does reveal something I've seen before that implies much about how charters get their "gains":

Algebra I is a course that you take in Grade 8 if you are relatively advanced in math (Grade 7 if you're really advanced). If math isn't necessarily your thing, you're much more likely to take it in Grade 9; in other words, high school. See what happened here? The CPS middle schools scored quite well on the test; better, in fact, than Passaic A&S if we take into consideration the variables in my model.

But Clifton HS students didn't do so well. Would anyone make the case that the teachers are worse at the high school, compared to the middle school, based on this? You'd be foolish if you did: Clearly, we are justified in thinking the difference between the middle and high school scores is due at least in part to student characteristics we can't see in the data.

Well, if that's the case, why shouldn't we at least ask the question whether the same thing is going on with the charter school? The students there are self-selecting into Passaic A&S; isn't it likely they are different from students who don't choose to go to the charter? Shouldn't we at least stop to consider that charter school "gains" are due, at least in some part, to self-selection?

Charter researchers have, in some cases, tried to get around this problem by using lotteries to claim they are setting up a "random" experiment that accounts for these differences. The problem is three-fold: first, the charter "treatment" might include things like free-riding on salaries, which allows for an extended day, which could lead to higher test scores. In this case, we aren't putting the public district schools and the charters on equal footing; in other words, we're confusing "charteriness" with advantages that are paid for by the hosting district schools.

Second, we aren't accounting for peers. I know some have tried to do so in their work, but I find their methods to be weak (I know this deserves more discussion -- this summer, I promise). The fact is that concentrating special needs students in the hosting public district schools likely gives charters an advantage that is, again, not due to their governance structure.

Third -- and this is the big one -- We can only generalize "random" charter school studies to students who enter charter school lotteries. Which is what I'm asking folks to consider with the chart above. If we can see self-selection play out this way in public district schools, why shouldn't we expect similar effects from charter schools?

Let me bring this back to the local level: 

Here's Clifton, NJ, a city school system that is being screwed over by the state, which refuses to follow its own law and give the schools what they need. That same state then forces the community to fork over millions of dollars to a charter school whose operations and governance are far from transparent. The charter doesn't take its fair share of special needs students; spends money on administration and plant costs rather than instruction and support; and free-rides on teacher salaries.

The Clifton BOE rightly turns to the state and says: this is not fair to the students we must educate, and it's not fair to our taxpayers who demand accountability and transparency. 

What's the state's response?
NJDOE spokesman Michael Yaple said the agency declines to address local issues through media outlets and instead attempts to work directly with municipal officials.
"The Department takes into account a number of factors when making decisions regarding charter schools, ranging from student academic performance to the demand among parents," Yaple said in a statement. "Throughout the process, any correspondence and concerns from both school districts and charter schools is important to the Department, and we work directly with local school officials the best we can."
In other words: the good people of Clifton can take a flying leap.

At some point, this madness must come to an end. Even if you are a supporter of charter schools, nothing can possibly justify New Jersey's totally screwed up system of charter approval, regulation, and funding. We can't keep asking cities like Clifton to sacrifice their local public schools just so a few folks can have a "choice" that is negatively affecting the entire system.

h/t the great Rob Tornoe.

* A 3-year average from 2012 to 2014; I haven't yet been able to get more recent data, but this, in my opinion, is a good enough metric for our purposes here.

** Percent black and percent Asian, which I've found avoids problems with multicollinearity.