Log in  |  Search

Programmed to Fail: The Parthenon Report and Closing Schools

In New York City, virtually all schools serve high need students. But some schools serve students with very high needs, and serve them in astounding concentrations that have been rising over the past few years.  That’s troubling enough, but what is more troubling is the prospect that the DoE implemented policies that led to the concentrations, even though DoE knew that such high concentrations were highly likely to overwhelm those schools. [1] Then when the schools were indeed overwhelmed, the DoE failed to give them the support required, labeled those schools failures, and then moved in to shut them down.

And that’s why the UFT has recently asked the  State Attorney General to determine whether or not DoE is failing to meet its legal obligation to provide “educational equity.”

A few details of the situation in New York are worth knowing.

Back in 2006, the DoE received a report it had commissioned, called the Parthenon Report. In it the researchers showed, clearly, that  concentrating students with high needs in a single school generally leads to dramatically lower graduation rates. In fact, the researchers actually predicted the graduation rate of different schools based upon the intensity of the concentration.  They also pointed out that the effect on concentrations was not limited to the influence they would have on at-risk students.  For example, the chances that an average student would graduate could swing 30 percentage points (from 55% to 85%) depending on the concentrations of high-need students, and to a much lesser extent, the size of the school.

No surprise to teachers, of course, but DoE needed a fancy formula to figure it out.

Yet having made this discovery, DoE policy people do not seem to have done much to address it.  Just the opposite: DoE seems to have created a web of policies that concentrated the kids most at risk in the schools they did not create.  For example, DoE excluded students who needed special classes (self-contained students) from some schools, which meant others had higher concentrations.  And it apparently sent high school students who showed up at the student placement office  (new to the country or the system, or in need of transfer from other schools or penal institutions) to older schools– even when there were empty seats in the newer local schools.  Sometimes, the newer schools were (and are) right in the same building.

These policies – and a high school selection process that allows new schools to favor students “known to the school” — seem to have a concomitant effect of concentrating at risk students and because of the concentration lowering their chances of success.

A few highlights of the results:

  • Brooklyn: one out of every three of Robeson’s students enter the school overage, but at the 9 new neighborhood schools that replaced Erasmus Hall and Prospect Heights only one in seven enter overage.  Yet, DoE documents show  the newer schools had more physical spaceand they also had seats available in most of them in most grades.
  • Bronx: at Christopher Columbus, the test scores of incoming students are in the bottom 10% of all City schools.  Two of the new schools located in the same building serve students with the highest scores in its district.
  • Queens: Jamaica, Beach Channel, Far Rockaway and Springfield Gardens are four high schools the DOE has closed or is planning to close in Queens. All have served large numbers of students who require self-contained classes. But 8 of the 9 new schools that are replacing them serve none of these students and the 9th serves only a small percentage. This includes two separate schools housed in right in the same buildings as Jamaica and Beach Channel High Schools.  In other words, in the 2009-2010 school year, even as Jamaica and Beach Channel worked with about 150 self-contained students between them, the schools co-located in their building served none at all.

There is hardly a public school in New York City that doesn’t serve some high need students. And in all of them – big or small, new or old – our teachers work with passion because that is the nature of teaching: the kids themselves and dynamics of the classroom inspire our work in ways that merit-pay advocates and teacher-haters have never understood. But the fact remains that the highest needs are concentrated in particular schools, and more and more that seems to be by design. Then, when the school “fails,” the DoE moves in to shut it down.  And, DoE is aided in that by the school Progress Reports  they created.   The Reports fail to account for key demographic factors and make differences in school results appear to have come about because of the Herculean efforts of some staffs, but not others.

The Parthenon Report has wide-ranging implications.  For example:

The odds of beating the odds are very slim. Of course, the real question is why the DoE would raise the odds in the first place when there were other options.   To put it another way, why build a mountain when you could have created a more level playing field?  But more importantly, there comes a point at which the mountain gets too high.  Parthenon – which is partly about how some schools beat the odds – defines perhaps 10% of schools as beat-the-odds schools. Yet a close look at the schools they highlight shows their needs are not as high as the schools that “fail.”

That doesn’t mean there aren’t great schools doing great things with challenging populations. Our international schools – which serve very recent immigrants– have programs that are by all accounts highly successful. But by the nature of their demographics, these schools are able to focus very specifically on a defined need.  And because they consistently beat the odds – there may very well be a socio-economic (demographic) factor at work within the recent-immigrant population that is unaccounted for in the usual formulas DoE uses to give us evidence of school success. Which brings us to the second point.

Most differences between school performance seem to be driven by factors that have been ignored. The “failing schools” movement (from Duncan to Klein, to a whole lot of people in between) makes its case for shutting some schools by exalting others. “These schools are similar,” they say, “and yet this one here has beaten the odds – and that one there has failed.” But once we account for the hidden demographic factors that influence success and are routinely excluded from accountability, (such as incoming, overage students) the picture becomes much different.

Take for example, the comments of DoE’s Marc Sternberg at a recent hearing at City Council. Sternberg compared closing schools to the six small schools on the Bronx Evander Childs campus where he said the average graduation rate was 80.3 percent.  What he did not point out was the hidden demographic differences between these schools and the typical closing schools, like Columbus High School down the block. Last year, the six Evander schools had average incoming scores of  2.7. The percent of self-contained students in these schools was  3%, and the percent entering overage for all grades was 15%. Sternberg particularly highlighted the school he led on the campus (Bronx Lab). That school served virtually no self contained students ( .2% ) only 11% overage, and average incoming scores of 2.8.   Compare all  this to Columbus   in the same neighborhood  where the average incoming scores are much lower (2.4) and the needs insanely higher (self contained 12%, overage 28%).

And it is not just Sternberg’s comments. Santi Taveras said pretty much the same thing last year, referencing other schools.  Even that 10% of schools designated as “beat the odd” schools in the Parthenon Report actually have different, and lower  concentrations of high-need students. Ultimately, it is extremely tricky  to draw any conclusions about differences in performance until we know for sure about the different levels of challenges our schools face.

The DoE must start talking honestly about the nature of the problem. Or to put it another way, if they know it, why not just say it?  Maybe Sternberg and Taveras were misinformed about their comparisons.  But how could they possibly be  blind to what their organization’s policies have wrought.   And even under the most benign of constructions – that it was all just a big and unintended mistake – we have to wonder why DoE has not been up front  about that big mistake.  DoE needs to say – honestly and forthrightly – that the odds are stacked against these schools, and that there is no good that can come of underestimating what these needs are.  Step one of recovery is admitting you have a problem, yet there is Cathie Black on NY1 telling the public that the only difference between closing schools and others is the “level of commitment” of the staff.   And while  Cathie Black couldn’t be expected to know about Parthenon, every one in her policy department certainly does.

Forget MDRC and lottery-in/lottery out research. Or, at the very least, read this kind of research with a different eye. DoE touts MDRC and other school-quality research that is designed around comparing students who apply through lottery-like admissions policies to small (and charter) schools.  Basically, they match similar students who get in with those who don’t and examine their academic progress.  If the accepted students have better outcomes (and that is not always the case) the researchers make claims about how the new school did a better job.  This research is then used – by DoE and others – to hasten the shuttering of schools.

Lottery in/out research is often considered the “gold standard” of school research.   What we see now, however, is that to the extent there are differences, a major driver of those differences is the effect of concentration on the school as a whole.  Instead of pointing to studies like the MDRC in order to justify their closing school strategy, they ought to take the actions that would be implied by differences, such as focusing real support and resources to the schools in which they have concentrated students[2], and focusing on the policies that led to concentration in the first place.

But that’s not happening. Between narrow ideologies and brutal politics, DoE has created a system that sinks some boats and raises others. But the “sinking” is not based on the failure of the captain and the crew, but rather manipulation of numbers, policies, and ultimately, our students’ lives.

Isn’t it time to look honestly at the challenges in these schools?  And raise all boats?

[1] To  teachers and principals the challenges brought about by concentrations of high need students are obvious.  Besides the significant influence of peer effect, there are the challenges of providing appropriate programming and support – especially given the limited resources with which they have to do it.  How does a school program a student who is hearing impaired and needs special “pull-out” services but who also has limited English  proficiency in English, and has limited literacy skills in his own language and can’t afford to miss a class.   Not as rare as you might think.  Ultimately, when many students have high needs, schools are simply overwhelmed.

[2] It’s worth pointing out, by the way, that MDRC itself partly attributes success at these schools to another DoE policy that protects these schools – reduced teacher load.



  • 1 Christine Rowland
    · Mar 10, 2011 at 8:33 am

    Great work Jackie! I attended a District 11 Town Hall Meeting with the Chancellor yesterday evening. In response to some young people from Brothers and Sisters United who spoke out against closing schools the Chancellor started to use Evander and their new small schools grad rates to show why the schools were being closed, and how their replacements were better. I really think she didn’t know the facts – she’s just been fed the official line.

    From the audience their were calls of ‘not the same kids’, and a network leader jumped in to take her place at the mike. The Chancellor was wisked out to another engagement (which has been announced at the beginning of the meeting).

    Using data to mislead happened again later in last night’s meeting after the Chancellor had left, as you report it happened at the City Council hearing. The problem was that this time it was by people who DO know better. Individuals who are in positions of authority and respect are knowingly not telling the truth about the populations in schools. These aren’t issues of opinion, they are facts.

    Is there no accountability anywhere? If this goes unchallenged the DoE will continue to systematically fail thousands of students in dozens of schools in this system each year.

    What is happening with the Schneiderman letter? Will there ever be justice for our communities?

  • 2 Remainders: The snow-day favor Klein paid Kennedy | GothamSchools
    · Mar 10, 2011 at 7:01 pm

    […] The Parthenon Report on school closures prompted the UFT to petition the attorney general. (Edwize) […]

  • 3 Online Education in America » Blog Archive » Remainders: The snow-day favor Klein paid Kennedy
    · Mar 10, 2011 at 7:06 pm

    […] The Parthenon Report on school closures prompted the UFT to petition the attorney general. (Edwize) […]

  • 4 Leonie Haimson
    · Mar 11, 2011 at 12:37 am

    This is a good analysis, but it should be taken into the courts. I don’t see what the AG is likely to be able to do about this. Another point: the International schools have tiny class sizes; 15 or less. And the small schools not only exclude our neediest students but offer smaller classes of about 5 students per class less as well. How do they do that? They can cap their enrollment at a lower level and get more space per student.

  • 5 Leonie Haimson
    · Mar 11, 2011 at 12:39 am

    Another point: the lottery method of comparing outcomes that ignores peer effects and the effect of concentration of high needs students in one site vs. another is a failing of the charter school literature as well.