I’ve been writing over the past few weeks about NCLB waivers and the schools that are being targeted by states under the waiver program as targets for federally endorsed state intervention. [all of which is built on highly suspect legal/governance assumptions]

My concerns here operate at a number of levels. First, the current Federal Administration has again used an “incentive” application process to coerce states to adopt really, really ill-conceived policy frameworks. These policy frameworks consist of two major parts:

1. school and district performance classification schemes that are largely if not entirely built on misinterpretation and misrepresentation of generally low quality data; and
2. poorly vetted, ill-conceived, aggressive/abrupt (closure, turnaround) intervention strategies as likely (if not more so) to do harm as they are to do any good.

So… yeah… it boils down to ramming bad, disruptive restructuring plans down the throats of schools/districts/communities that have been classified by biased, and unjustifiable measures. Further, much of this is being proposed without carefully evaluating whether there exists legal authority to do any of it.

Junk Classifications 101

So, let’s take a look at how the school classifications have played out in New Jersey. New Jersey, like other states proposed to classify its worst schools as Priority schools – subject to immediate disruptive intervention, the next lowest set as Focus schools – the you’re next/we’re watching you schools – and another set as “reward” schools – or you kick ass so we’re gonna give you a prize!

Matt Di Carlo over at Shanker Blog has given considerable attention to the issue of state school grading systems and the extent to which they measure or even attempt to measure school effects on student test scores (not to be conflated with actual school “effectiveness”), or instead simply capture the compounded influence of a variety of student background factors on various accountability measures. In other words, are school ratings simply classifying poor minority schools as bad schools and thus branding their teachers and administrators as necessarily ineffective, while not even attempting to actually discern their effectiveness.

Further, in my last post on New York City schools I showed that while there were subtle differences in mean teacher percentile rank across schools rated as the worst (priority) versus those rated best (good standing), a) there were still many “best” schools where teacher average test score effect was much lower than in “worst” schools and b) schools that had lower income students and more minority students were still much more likely to be rated as among the “worst” even if their teacher “effects” were similar.

New Jersey Classifications

This first figure shows the demographic composition of schools by their classification. Perhaps the most astounding feature of this graph is that priority schools are nearly 100% black and Hispanic, while reward schools have very low levels of low income, black or Hispanic students.

Here are a few maps to illustrate the geographic distribution of priority, focus and reward schools, for those who know Jersey. We can see that the priority schools are concentrated in the larger, poorest urban centers and focus schools in and around other poor cities/towns.

Not surprisingly, the reward schools for the most part are scattered through the more affluent suburbs of northern Bergen County and out through the most affluent areas of north central NJ (Morris/Somerset/Hunterdon). Okay… I was actually surprised that they had concocted a rating system that was so absurdly biased. The second set of maps shows that there are some reward schools in the northern half of the city of Newark (the area with lower black population share).

Underlying Measures for Classification

It was assumed that states would be proposing ratings based on a mix of status and improvement measures… and that doing so would somehow mitigate the extent of demographic bias in the classifications. States could also use subgroup and achievement gap measures. States wouldn’t, for example, simply be proposing to step in and close down all of the majority low income and minority schools and turn them over to private management/or otherwise displace their entire teaching and administrative staffs.

Of course, the measures available in most states aren’t always that useful to sifting through the demographic biases.New Jersey’s are particularly bad. The following figure shows the racial and low income composition of schools by the types of measures that determined their status. Both the progress ratings and the performance level ratings are hugely biased! As it turns out, so are the achievement gap and subgroup measures. Notably, many affluent New Jersey districts (where the reward schools are) likely have too few low income or minority students to even report gaps.

Remedying Poverty by Deprivation?

In my analysis of New York State, I also showed that priority schools are far more likely to appear in school districts that have been most underfunded by the state of New York relative to its own promised school funding formula (the one the state adopted/proposed as a remedy to court order several years back).

Now, New York state has one of the worst state school finance systems in the nation. One in which districts with more needy students have systematically fewer resources. New Jersey is a far cry from New York in this regard. New Jersey has done better than most states with respect to funding equity and adequacy. 

And compared to demographically similar states, New Jersey has some positive results to show for its overall funding effort and for its targeting to high poverty districts.

But lately, New Jersey has started down a different road in state school finance policy. The state has chosen in recent and proposed for future years to significantly underfund their own legislatively adopted state school finance formula.

That in mind, the following slides present an analysis somewhat similar to that presented in New York State, but looking forward instead of back. I’m not proposing some lofty “what should be” funding levels based on academic analysis here. Rather, I’m simply looking at the extent to which New Jersey is currently, and proposed to fund districts under its own formula SFRA. This is the formula that was adopted by the legislature under the previous administration and was subsequently upheld by the state court. More on these issues in a later post.

I’ve not had time to reconstruct my own simulations of SFRA projected out over the next several years, so I’ve used data pulled together by the Education Law Center and SOS NJ in which they have projected (SOS NJ) out the SFRA funding shortfalls for each district for the next 5 years. The figure below shows that in the current year, funding shortfalls from the current legislated formula are smaller in districts that are home to priority and focus schools (note that the formula itself significantly reduced targeted effort to these districts when it was implemented).

But, over the next few years, it is expected that as these schools – priority and focus – are subjected to takeover/overhaul/closure – their districts will be increasingly shorted in their funding with respect to what the formula estimates.  That is, the overall strategy here appears to be to identify high need schools for takeover/closure and then systematically and substantially reduce their financial support over time.

Cumulatively, over the next five years, districts of priority schools stand to lose much more on a per pupil basis (relative to what the formula dictates they should receive) than districts of reward schools.

Put bluntly, the goal is to “reform”(?) priority and focus schools and close achievement gaps by taking all of that harmful money away from them and giving it to others who are far less needy! Yeah… that’ll learn-’em!

This is all strangely consistent with the framing of the commissioners report, that was not a report, on school funding and achievement gaps in New Jersey. In that report, Commissioner Cerf essentially proposed (via a series of bad and worse graphs) that the road toward closing New Jersey’s achievement gap should be paved by reducing funding to high need minority districts and shifting it to lower need, lower minority concentration districts. Strange logic indeed.

And these reductions presented above don’t account fully for the plethora of other alterations proposed to the state school funding formula that might further reduce funding to higher need districts – funding to districts that are home to priority and focus schools.

The following posts critique some of the proposed changes, and address other related issues:

• How student enrollment counts matter in school funding
• Stuff that actually needs fixing in SFRA
• Are free lunch counts a reasonable measure of educational “risk?”
• Does NJ really have a large achievement gap relative to other states?

Closing Thoughts

As I noted on my previous post, I can hear the reformy outcry now that this is all warranted because we’ve provided poor and minority kids the worst schools and worst teachers for so many years. This is merely an attempt to remedy this persistent, intractable disparity.  The problem with this logic is the placement of blame (in addition to the questionable legal authority and ill-conceived remedies).

We’re not measuring school performance here. There’s no basis in these classification schemes for implying that the teachers and administration are the ones who failed the children. These are junk, gerrymandered classification schemes. They are based on arbitrary distinctions being made with inadequate data/information.