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

Sunday, September 14, 2014

Once Again, @tomamoran Gets Charter Schools Wrong

So long as Tom Moran, Editorial Page Editor of the Star-Ledger, insists on publishing pieces about education full of omissions and half-truths, I have no choice but to continue to set the record straight and correct him.

Today's piece from Moran is about Hoboken Dual Language Charter School (HoLa), a dual-language immersion school I have studied previously and know quite well. Says Moran:
The core dispute is about race. Whites in Hoboken have fled the district in droves, thanks to its long record of academic failure and racial imbalance. The city is 82 percent white, but all the district elementary schools are majority African-American and Latino.
In other words: whites have fled Hoboken's public schools because whites have fled Hoboken's public schools. Nice insight, Tom...

We'll get to whether the "core dispute is about race" in a minute. But let's first take a look at racial makeup of Hoboken's public schools:

There is no doubt that HoLa and the other Hoboken charters have different student racial profiles than the Hoboken Public Schools. So when Moran makes a nasty insinuation about HPS parents:
And the dirty secret is that the district itself is aggravating segregation by allowing white families who live near Connors to travel across town and enroll in other schools. 
The mechanism is a perverse district choice program. In Montclair, parents rank their choices and are enrolled with the goal of achieving racial balance. In Hoboken, choice allows white families to flee from Connors, making segregation worse.
He's not giving the full story. No Hoboken school besides Brandt -- a pre-K/K school and, therefore, not really a relevant comparison -- has as high a proportion of white students as HoLa. If the white parents whose children are in HPS schools are "fleeing" Connors, the same can easily be said of the parents of the charter schools.

I've been very careful not to make these sorts of accusations about Hoboken parents, whether they send their children to HPS or the charters. I guess Tom, a professional journalist, has no such qualms.

And he appears to have little interest in addressing the disparities above directly:
And in the end, they [HoLa] got twice the portion of minority kids as the city’s population. And progress continues. This year’s kindergarten class is 41 percent minority. 
So ask yourself: Is this the profile of a school that is trying to block out minority kids?
First of all, define "minority." If "minority" equals "non-white," then the kindergarten class is carrying on a tradition of HoLa having a 60% white population, which is significantly different from all of the HPS schools.

Second: as I explained in my post about the socio-economic segregation found in Hoboken's charter schools, you should not compare the demographics of school-aged children to the demographics of a population that includes adults. I was just in Hoboken last night, and as anyone who has been there can tell you, it is a mecca for young urban hipsters, many of whom don't have children.

Assuming that you can extrapolate the racial composition of the student population (public, private, and home schooled) of Hoboken by simply looking at the racial composition of all age groups is a rookie mistake. School-aged children only account for 9 percent of the total population of Hoboken -- their racial makeup could be quite different from the entire population's.

I don't have cross-tabs on age and race for the city (I'll keep digging...). But when I looked at age and socio-economic status before, it was clear that judging Hoboken's charters against the total population of Hoboken was bound to lead to false conclusions:

So there's every reason to doubt a comparison between the racial profile of the entire city population -- including adults -- and the racial profile of the Hoboken charters. Unless Moran has data I don't have that includes all of the children of Hoboken without adults, he ought not to make these comparisons.

Further: Moran was the one who said race was the "core issue." He didn't care to address other issues in segregation that may be far more germane to this discussion:

Here are the Free Lunch and Free & Reduced-Price Lunch eligibility numbers for Hoboken's charters. There's no doubt that Connors has many more students in economic disadvantage than HPS's other schools. But there is also no doubt Hoboken's charters educate far fewer students in economic disadvantage than any public school in Hoboken (save, again, Brandt).

And it doesn't even end there:

Hoboken has a relatively low Limited English Proficient population; that said, HoLa didn't educate one child who was listed as LEP in 2013-14. 

I had to go back to 2012-13 for special education data: NJDOE reports HoLa had 0.0% special education students that year, and 0.6% in 2011-12. This is school-level data*; district data puts HoLa at 4.47% (honestly, I don't know why there is a disparity, especially because NJ considers a charter school to be its own district). 

You might look at this and think: "Well, HoLa isn't taking it's fair share of kids with special needs, but the other charters are doing OK."

Not quite:

An SLD is a "specific learning disability." SPL is a speech or language impairment. These eligibilities, while certainly deserving of services, are the less costly categories of placements. The plain truth is that HPS is not only taking more children with special education needs than their neighboring charters; they are taking more children with the costlier needs. You simply can't make the relative comparisons that Moran does about charter school budgets without addressing these fundamental facts.

When Moran says race is the "core issue," not only is he providing an incorrect context for assessing the state of Hoboken's charters -- he's ignoring issues that are at least as "core" to the charter school discussion as race. He's not addressing disparities in economic disadvantage, and he's not addressing disparities in student needs. As Bruce Baker has pointed out, Hoboken's charter sector displays some of the greatest economic disproportionality of any community in New Jersey, let alone the country. Where is Moran's concern about this?

One more thing:
And their [HoLa's] test scores are rocking: They are in the top quarter of academic achievement in New Jersey, according to the state, and the top 1 percent when measured against peers of the same demographic. 
So, what, exactly is the problem?
As Julia Sass Rubin has pointed out, the state's "Performance Reports" are highly questionable measures of student achievement; the precise methodology for creating the "peer groups" remains hidden, so we don't really know how it works.

Using public data, I looked at the achievement of HoLa and other Hudson County charter schools, taking into account student characteristics. The results weren't "rocking" -- in fact, they were precisely what all the research on student poverty and achievement have led us to expect:

When you compare HoLa to the rest of Hudson County's schools, you find that its proficiency rates are below what you would expect for a school with so few students in poverty. The rest of the county's charters don't do much better. In fact, many traditional public schools get results just as good or better than HoLa, yet serve many more students in economic disadvantage.

From my report:
Certainly, there is no evidence within the NJDOE data to show that charters in Hoboken and Jersey City are engaging in a deliberate pattern of cream-skimming. That same data, however, is quite clear: the charter schools in Hudson County that have higher rates of proficiency and/or student growth do not serve the same percentage of economically disadvantaged students as their neighboring traditional public schools.
Further, there is a clear correlation between these charter schools’ test outcomes and the relative percentage of free lunch students they serve compared to their neighboring TPSs. The correlation is much stronger for the county’s charter schools than its TPSs.
Hudson County’s policy makers, education leaders, and citizens need to ask themselves a question:
Are cream-skimming charters a good investment if their test score outcomes correlate closely with their disparity in serving economically disadvantaged children?
This is, of course, a question Tom Moran will never ask. He'll fret about "tone" and he'll accuse people like me of saying things we never said. But he will never, ever ask himself the hard questions about education policy in New Jersey.

How sad.

The Star-Ledger Editorial Board meets, once again, to discuss education.

ADDING: Just saw Moran posted this in his comments:
To those who press the demographic contrast between hola and district: No one disputes there is an imbalance that's a problem. The question is what the answer is.

Wouldn't it make sense to allow Hola to use preferences in its lotteries? And wouldn't it be a travesty to halt the expansion of a school that is so obviously doing well by its kids, one that hundreds of Hoboken families want to enroll in?

We've been down this road a million times, Tom. I know people way smarter than me have tried to explain this to you. Yet you just don't get it:

If HoLa or any other charter is "obviously doing well by its kids" by serving a different student population, it logically follows that what they are doing is not replicable on a larger scale.

You can't make every school in Hoboken have a proportion of its students in economic disadvantage that is less than average -- this ain't Lake Wobegon. Somebody has to educate the free lunch kids. Somebody has to educate the LEP kids. Somebody has to educate the special eduction kids, especially the ones with the most costly needs.

You say you want to put more of those kids into the charters. But all indications are that when you do, the charters will not have performed as well as the public schools. Look at this again -- seriously, really look at it.

I used one aggregate measure here; we could use a bunch more, but I'm telling you we'd pretty much see the same thing. Yes, HoLa has a decent aggregate proficiency rate; good for them! But they serve far fewer kids in economic disadvantage than the other schools. Their outcomes are actually below what we'd predict. I would never say they are a "bad" school for not precisely following this trend. But I will point out there are many public schools that do as well or better than HoLa on test-based outcomes even though they serve a much greater proportion of at-risk children.

This is not rocket science; if it were, I couldn't do it. This is really, really basic stuff. Why don't you understand it?


* I have this data as part of an OPRA request; it isn't available publicly in a database, even though you can find it in the school performance reports. Weird, huh?


Veri said...

I don't understand aren't hispanic children already dual language learners so if they were to go to HOLA they would actually already have been immersed in two languages, would the school then have to choose a language other than spanish or portuguese to level the playing field? I find it kind of ironic that people are fleeing the school with the children who speak dual languages to go the charter school where they are purposely immersed in two languages. ...as my students would say...Just Sayin.

Marie Corfield said...

Not all Hispanic children have "already been immersed in two languages." Many enter school not speaking English at all. They have to learn English while learning other subjects.

Marie Corfield said...

... and some enter school without a basic command of either language. Huge hurdle to overcome.

Teacher Mom said...

THat was a head scratcher for me too. How can a "dual language" school NOT have a single LEP student? Then what the hell is the purpose of a "dual language" school?

CrunchyMama said...

@Teacher Mom: my younger child is in a language immersion program in our MD school district; while there *are* LEP students in the program (mostly those who's target-language-speaking parents have enrolled them there), that's not a specific feature of the program; the majority of the kids there are native English speakers, and the same would likely hold true if it were a dual-language program, the point of which would be to develop limited proficiency in a target language while keeping some instruction in (presumably) English.