Tuesday, July 5, 2016

Abbott Spending Has Been Ineffective

Hey Education Law Center
& NJ Supreme Court!

I am completely against Chris Christie's proposal to give equal state aid to economically and demographically unequal school districts. Christie's "Fairness Formula" is neither "fair" nor a "formula" and would devastate poor districts throughout New Jersey and ultimately the towns themselves, as they attempt to sustain school spending by taxing themselves to death.

 I'm an adult and I'm used to BS arguments from politicians, but I had a visceral reaction to Christie's attempt to cloak an inequitable idea in equity in his claim that he was "valuing each child and their hopes, dreams and potential the same."

And yet, just because Christie is wrong about the Big Idea of his proposal doesn't mean he is wrong about absolutely everything.

Chris Christie, was right about two things which I hope to explore in this post and a subsequent post.

1.  The NJ Supreme Court "overcorrected" for disparities that existed in the 1980s and Abbott Adequacy targets are beyond the point of effectiveness.

2.  The Abbott spending mandate is one reason, of several, for New Jersey's property tax crisis.

This post is going to be about the marginal effectiveness of "Abbott Remedy." Please note, I'm not saying that the Abbotts and other poor districts don't need superior state aid.  I'm saying that the "Parity Plus Doctrine" and the high Adequacy Budgets for high FRL districts embedded in SFRA are higher than will be effective.

I consider the "money doesn't matter" argument that Chris Christie is making to be ludicrous.  I am neutral on charter schools.  If New Jersey had a flourishing economy I'd want to fully fund SFRA.  I believe that poor districts must receive above-average state aid to make up for their insufficient tax bases and to sustain opportunity for their children.  However, I'm strongly anti-Abbott for three major reasons.  First, I believe it is unfair that the Supreme Court's "Parity Plus Remedy" applied only to an arbitrarily developed list of urban poor districts (later known as the Abbotts) and not equally poor districts who for obscure reasons were never put on the Abbott list.  Second, the amount of money it directs to poor districts is excessive.  Third, the State of New Jersey is increasingly insolvent and what it could afford in the 1990s it cannot afford today.

The Scale of the Abbott State Aid and Spending Advantages

Unfortunately, too few people make the rational comparison of Abbott districts to poor non-Abbotts.

Chris Christie himself missed the relevant comparison this when he introduced his "Fairness Formula."  He introduced his "Fairness Formula" at Hillsborough High School, a very affluent district, and he compared Abbott academics to districts that are nothing like them.

Just take a sample of graduation rates. The statewide graduation rate is 90%. How have we done in the 31 districts where we have invested $97 billion over the last 30 years? Asbury Park—66%. Camden—63%. New Brunswick—68%. Newark—69%. Trenton—68%. 27 of the 31 districts are below the state average, despite the exorbitant spending over the last 30 years. There are exceptions and those should be noted right here. In Harrison, Long Branch, Millville and Pemberton they have exceeded the statewide graduation rate. In Union City, the have seen extraordinary growth under very trying circumstances and the leadership in those districts deserve great credit.
Comparing the Abbotts to the state as a whole is idiotic. The Abbotts are 77% FRL-eligible and the state is 38% FRL-eligible (and that includes the Abbotts), so of course the Abbotts are going to lag the state academically. And yet the Abbott/statewide comparison wasn't the only bad comparison that Christie made, because he also compared the Abbott district-run schools to a handful of the highest-performing Abbott charter schools, despite the well-known fact that charter schools have non-representative enrollment.

Many people who are in favor of Abbott compare the Abbott districts to poor districts in other states or make a chronological comparison of the Abbotts now to the Abbotts in the past.  Some people will say that Abbotts are making progress as if that were evidence of Abbott Funding efficacy and ignore the parallel progress poor non-Abbotts are making too.

Although SFRA was written for the benefit of poor and working class non-Abbotts, journalistically and in blogs, low-income non-Abbotts are the forgotten districts of New Jersey.  Although there are 78 districts in DFGs A and B who are not Abbotts and 64 non-Abbotts in DFG CD, these districts get very little attention and Christie has ignored them, just like the NJ Supreme Court has. Chris Christie spoke of NJ's heavy tax burden, and yet he didn't mention the poor non-Abbotts where the tax burden is the highest, services usually the worst, and where housing values are falling.

The Abbott/low-income non-Abbott comparison should be first in any objective person's mind because the state aid advantages of the Abbotts are tremendous and thus New Jersey presents a judicial experiment in what the effects of higher school spending are.

The disparities in state aid and spending between Abbotts and poor non-Abbotts are very stark. Of the 37 districts in NJ who have Local Fair Shares below $5,000 per student, 20 are Abbotts and they get $15,600 per student, on average. The 17 non-Abbotts among the poorest only get $10,400.

The Abbotts also get two years of universal Pre-K whereas poor non-Abbotts get nothing. Although not all Abbott facilities are superb, there are more finely built, state-funded new buildings there than there are in poor non-Abbotts.

Yet the Abbott spending advantage is less than the Abbott state aid advantage because 29 out of the 31 Abbotts tax below their Local Fair Share and 14 have taxes below even 50% of their Local Fair Share. Low-income non-Abbotts do not have the luxury of undertaxing and usually accept very high tax burdens. A handful of Jersey Shore resort towns and Atlantic City have high tax bases and have high spending, a handful (like Lawnside) get high state aid despite not being Abbotts.  Nonetheless, most poor non-Abbotts do not have decent tax bases and most do not get high state aid.

Even after SFRA officially ended the "Parity Plus" regime, the flat state aid of the Christie Era, and the Abbotts' own refusal/inability to pay school taxes, the 31 Abbott districts have an average (unweighted) Total Budgetary Cost Per pupil of $17,961 for 2015-16. The 31 demographically poorest non-Abbotts have an average Total Budgetary Cost Per Pupil of $14,067. (see this spreadsheet for more spending information.)

Some of the poorest non-Abbotts are K-8s and their average FRL-eligibility is slightly lower, so somewhat lower spending is to be expected, but the gap should not be of this magnitude.  Even not counting their PreK money and state construction money, the Abbotts have a 28% financial advantage in opex aid.  

Abbott Results?

There are some involved in the Abbott Cause for whom the effectiveness of Abbott has always been secondary because they because they believe that equal opportunity is an end in and of itself and without high state aid and high spending, equal opportunity cannot exist.

The goal of Equal Opportunity was how Education Law Center lawyer Marilyn Morheuser parried the argument that high spending would be effective. "The state can't guarantee results. It can guarantee opportunity." Chief Justice Robert Wilentz also deflected the argument that money would not work by saying that the NJ Constitution still required high spending in urban poor districts, anyway whether the kids would benefit or not, "Our constitutional mandate does not allow us to consign poorer children permanently to an inferior education on the theory that they [sic] cannot afford a better one or that they would not benefit from it."

Gordon MacInness, now head of the New Jersey Policy Perpective and formerly the DOE administrator in charge of Abbott, said in 2009 at Teachers College that Abbott was worthy because it narrowed the gap in "life chances," ie, opportunity, but admitted in terms of long term academic results Abbott was a "huge failure."

Gordon MacInnes, a fellow at the Century Foundation who oversaw implementation of theAbbott decisions as Assistant Commissioner at the New Jersey Department of Education from 2002 to 2007, delivered a mixed assessment. The gap in “life chances” between poor and middle-class and wealthy students in New Jersey and across the nation is “still substantial,” he said. “When you get to middle school, eighth grade, high school – forget about it. This has been a huge failure.” 
MacInnes said Abbott, “arguably the most important judicial decision in education since Brown v. Board of Education” in 1954, has resulted in some meaningful changes in individual school districts for certain groups, such as free, high-quality preschools for three- and four-year olds. And some Abbott districts such as Union City, which instituted its own intensive K-4 literacy program and got adequate support for teachers, have done exceptionally well. But it has not produced the hard, concerted effort between students, teachers and parents – and, especially, the support of teachers of low-income students – that is necessary to get poor students on the same footing with wealthier ones.

Nonetheless, there are some researchers who are interested in whether or not Abbott spending has produced academic progress.  Although there are some researchers who claim that Abbott spending has produced large results, others admit that the evidence is contradictory and the gains that have occurred in Abbotts occurred in low-income non-Abbotts too.

The following chart is from a PhD dissertation by Joanne Howard (not online).  It shows that the Abbotts have gained more in percentage terms than low-income non-Abbotts, but that they started from a lower position, so they had more room to improve.

If you examine the chart above you will see that the Abbotts and "other low-income districts" made nearly parallel gains.  The wealthiest districts hit ceiling effects and did not gain.

Another researcher (Alexandra Resch, page 104) who concluded that the Abbotts did gain more than other districts nonetheless found much evidence of parallel gains in other districts.  (Here the comparison group is the entire state, not exclusively poor non-Abbotts)

By high school the Abbotts indeed do very poorly, as I will show below.

Yet the above charts are incomplete because they do not show what the correlation coefficient is.  I.e., they don't show the tightness of the trendline.  Yes there is a general upward trend, but you have no idea what the "noise" is.

If you separated out all the districts, you might have something like this from the Peak Abbott years of 2002-2007, which is what Chris Cerf found in his 2012 Education Funding Report.

Click to Enlarge.  Source, Education Funding Report:
What about today and contemporary academic position relative to poor non-Abbotts?

Abbott spending remains very high compared to poor non-Abbotts, but since the state's Adequacy targets are so high, all but six of the Abbotts are technically below Adequacy. (Based on low definition of spending/high definition of Adequacy model used by the Education Law Center.)

And yet, if we just look at the six Abbotts whose spending still exceeds their Adequacy Budgets and use the Peer Comparison Group percentiles that the Department of Education produces, we see no support for the contention that Abbottists have made that there is a strong correlation between high spending and academic achievement.

If the Abbott Hypothesis were correct, the highest spending Abbotts would outperform their peers and fiscally crippled poor-non-Abbotts would lag their peers.  And yet, when you actually look up the data, the low-spending districts actually outperform the high-spending Abbotts.  

For 2013-14 the DOE compiled an overall Peer and State comparison percentiles that were based on all of the standardized tests taken by students at the school.

In 2014-15 the state did not produce these comprehensive comparisons due to widespread PARCC opt-outs.  Since I do not have as good data for 2014-15, I will rely on 2013-14 data although I will parenthetically provide SAT data from 2014-15.  The Adequacy spending data is from 2015-16.  I admit it is for two years after, but I assume based on flat-funding that the differences in Adequacy Budget would be small over the two years.  If anything, the Abbotts were more above Adequacy then.
  • Asbury Park is $5,617 per student above Adequacy.  In peer performance it was at the 8th percentile.  In state performance it is at the 3rd percentile.  (Its 2014-15 SAT scores it is at the 19th percentile in peer performance.
  • Hoboken is $2,219 per student above Adequacy.  Its peer performance was at the 52nd percenIn state performance it was at the 18th.  (Its 2014-15 SAT scores are at the 52nd percentile.
  • Pemberton is $2,866 per student Above Adequacy.  In peer performance it is at the 31st percentile.  In state performance it is at the 16th.  (Its SAT scores are at the 65th percentile.  Pemberton's peer group is almost entirely non-Abbott and again, the district is 50% white.)
  • Phillipsburg is $981 per student Above Adequacy.  In peer performance it was at the 16th percentile.  In state performance it is at the 38th.  (Its SAT scores are at the 100th percentile, though this seems like an outlier among Phillipsburg's other scores) Phillipsburg High School is 72% white and only 29.8% Economically Disadvantaged.  It should not be an Abbott.
  • Keansburg is $3,475 per student above Adequacy.  In peer performance it was at the 72nd percentile.  In state performance it was at the 16th.  Its SAT scores are at the 84th percentile in peer performance.  Keansburg's peer group is all Abbott except for Manchester Regional and Lakewood, two of NJ's most financially crippled districts.  Also, Keansburg is not typical of the Abbotts demographically because it is 64% white and only 1.1% ESL.  The schools it is compared to have far more African-American students.  Keansburg's strong performance is more an example of the white-black achievement gap as it is of high-spending driving success.
Were I to include Pleasantville and Gloucester City, both of whom were nearly at Adequacy, the same pattern would continue.  

And now let us look at the Academic Achievement of the most financially disadvantaged non-Abbotts in New Jersey.  The schools below all have FRL-eligibility rates above 70%. Demographically these districts are just as poor as the Abbotts and yet spend $10,000-$12,000 per student.

  • East Newark is $10,097 per student below Adequacy.  Its one school performs at the 52nd percentile in peer Academic Achievement and the 20th percentile in state Academic Achievement.  It is so underbudgeted it cannot offer algebra.  
  • Fairview is $7,906 per student below Adequacy.  Its middle school performs at the 67th percentile in peer Academic Achievement and the 29th percentile in state Academic Achievement.  It is also so underbudgeted it cannot offer algebra.    
  • Bayonne is $7,307 per student below Adequacy.  Its high school performs at the 82nd percentile in peer Academic Achievement.  It is at the 40th percentile in state Academic Achievement.  (Bayonne's middle schools are also above average Ex 1, Ex 2, Ex 3)
  • Prospect Park is $7,087 per student below Adequacy.  Its K-8 school is at the 53rd percentile in peer Academic Achievement and the 19th percentile in state academic achievement.  It is also so underbudgeted it cannot offer algebra.
  • Guttenberg is $8,384 per student below Adequacy.  Its K-8 school is at the 64th percentile in peer Academic Achievement and the 29th percentile in statewide Academic Achievement.  It also does not offer algebra.  
  • Freehold Boro is $7,061per student below Adequacy and has the state's worst crowding.  Its middle school is at the 38th percentile in Academic Achievement and the 22nd percentile in statewide academic achievement.

  • Red Bank Boro is $6,759 per student below Adequacy and has the highest percentage of students who are LEP classified in NJ. Its middle school is at the 80th percentile in peer Academic Achievement and the 19th percentile in statewide Academic Achievement. 

There are studies out there who show that the Abbotts as a whole or some of the Abbotts outperform demographic peer schools who spend less money, but academic comparisons of budgetarily advantaged Abbotts with the most disadvantaged non-Abbotts should give defenders of the high Adequacy budgets some pause.

I compared some unlike schools above since I looked at Abbott high schools and non-Abbott K-8 schools (except Bayonne).  Lest anyone say that this is not a sound basis to conclude that Abbott spending levels are ineffective, I'm also including two graphs of Grade 8 ELA and math scores from 2013-14, plus SAT scores, plus AP/IB course taking.  As you can see, the severely under Adequacy non-Abbotts outperform the Abbotts.

(Note, Oresko school in Bayonne houses Bayonne's G&T program, hence the high scores, but even Bayonne's other middle schools outperform the over-Adequacy Abbotts.)

Finally, under-Adequacy Non-Abbotts outperform over-Adequacy non-Abbotts in the percentage of students taking at least one AP or IB test.

Note, there was an obvious data error for Belleville in the School Performance Reports. I used the percentage of Belleville students taking an AP test, not taking an AP class.
The gap between the Abbotts and under-Adequacy non-Abbotts in test taking, as opposed to course taking, is not as wide, but I believe that more of the above-Adequacy Abbotts pay for students to take the exam.  AP/IB test taking rates are thus not a fair comparison IMO.

Union City versus Dover

Union City is always brought up as the "Abbott that Works."  Liberals praise it and so does Chris Christie.  There is even a book about Union City, "Improbable Scholars."

Yes, Union City is a very good school system.  Its students are 94% Economically Disadvantaged and 13% Limited English and yet its students have a high graduation rate and perform well on state tests.  Its students are not consistently above average statewide per se, but they outperform their demographic peers and are in the middle third of the state on some tests.

And yet, Union City is a very poor argument for the effectiveness of Abbott.

First, whenever you have 31 of anything, at least one of them is going to be better than the rest. There are 31 Abbotts and statistically speaking, at least one of them is going to be very good.  Union City by itself is no more representative of Abbott effectiveness than Asbury Park is representative of Abbott failure.

Second, while Union City spends $17,500 per student and is above the state's average, it is more than $5,000 per student below Adequacy (due to underaiding and undertaxing), so Union City actually doesn't even spend at the level Abbottists claim is necessary for a district with Union City's demographics.

Third, Union City is not clear-cut as New Jersey's highest performing high-FRL district.

The highest performing high-FRL district in NJ is Dover.   

Dover's FRL-eligibility is about 66%, so it is not as high as Union City.  Dover is 9% FRL-eligible
too, which is lower than Union City's.
Union City HS

Not to disrespect Union City, but Dover's schools are higher performing.  30% of Dover HS students exceed 1550 on the SAT, 10% of Union City's do.  On statewide averages from 2013-14, Dover outperforms 65% of NJ schools, whereas Union City outperforms 24%.  35% of Dover students take at least one AP course, only 13.7% of Union City HS's do.

Dover Middle School is over 80% FRL-eligible and is a peer of the Abbotts demographically. On 8th grade math, Dover is at the 100th percentile in peer performance and the 69th percentile in statewide performance.

On the 8th grade Science exam, 75% of Dover students are Advanced Proficient or Proficient, a figure nearly equalling Montclair.

Dover HS
And Dover's spending is only $10,600 per student.  Union City spends $7,000 per student more than that.

Dover's higher performance is to be predicted based on demographics, but Dover exceeds its peer group by a wider margin than Union City does.

Why is Dover so good?  I have NO IDEA.  I've never read a single article about it.  Dover is just another example of how the media, researchers, blogs, and politicians ignore poor non-Abbotts.  Dover is a national Blue Ribbon school, but compared to charter schools and Union City, it has never gotten any press about it.

Does Dover need more money?  Absolutely yes, but Dover is evidence that the Abbott Dictum "money is necessary, not sufficient" is not consistently correct.


Some Abbottists look at particular studies and say that there has been an Abbott Effect.  Some admit that by high school that Abbott hasn't worked.  Some, like Education Law Center executive director David Sciarra in 2010, said that the data isn't there either way:

“There is a thinness of data over time,” said David Sciarra, of the Education Law Center, which represents the Abbott districts. He said evaluation systems that should have been set up a decade ago never developed. He is concerned that the new funding formula [SFRA] could make the situation worse. 
“There is no real assessment of whether the money is paying off,” Sciarra said. “The Rutgers study never came out.”... 
Sciarra said some Abbott schools have thrived, while others still perform miserably. Better evaluation is needed to hold all schools accountable. 
“We need a lot more research,” Sciarra said. “How long can they just keep blaming the Abbotts for everything?”

And yet, for the sake of argument, even if we accept that there is an Abbott Effect and that money is decisive for poor student population, where does that leave us regarding high-need non-Abbotts?  If money is so decisive then, how long can New Jersey deny sufficient state aid to equally poor and working class non-Abbotts?

And by what budgetary scenario can New Jersey fully fund SFRA while fully funding pensions?

NJ is putting $1.8 billion into pensions for FY2017.  That's a record amount, but over $3 billion less than the actuarial payment.  Yes, New Jersey must raise taxes, but raising taxes enough to fully fund pensions, pay for other state obligations, and fully fund SFRA isn't realistic.  If any person says "millionaires have gained the most in the last fifteen years, let them pay" then they should talk with Gov. Dannel Malloy of Connecticut and ask him if he thinks that tax rates have no influence on business location decisions.

And there will always be another recession too, during which even the $1.8 billion pension payment and $8 billion K-12 aid stream will be endangered.

I've said repeatedly on this blog that Adjustment Aid must be eliminated.  I've said that NJ needs to cut off all aid to ultra-high tax base districts.  I've said that Interdistrict Choice must be reformed.

But the problem is, that eliminating Adjustment Aid, eliminating aid to ultra-high tax base districts, and even halving Interdistrict Choice money only brings NJ to about $600 million.  Fully funding K-12 SFRA (with redistribution) would cost $1.5 billion.  To fairly funding middle-income, working class, and poor non-Abbott districts, we have to lower the Adequacy targets for high-FRL districts. Although this would have a hypothetical effect on poor-non-Abbotts, since non-Abbotts are drastically underaided, in reality, it would only affect the Abbotts.

Democrats have depicted attempts to lower Adequacy weights for at-risk students as a regressive idea and twice rejected proposals to do this (in 2012 and 2016), but when NJ cannot meet financially meet the existing high-Adequacy tarets and scores of districts are severely underaided, the high-Adequacy targets seem unjustifiable to me.

When the Abbott II decision came out in 1990 NJ was America's richest state.  We had little debt and a AAA bond rating.  We had had decades of solid economic growth.

Now NJ still has a very high per capita income, but it is closer to the national mean than it was in 1990.  Our bond rating is one of the worst in the country.  Since 2001 our economy has grown at half of the national average.

Even if you believe that Abbott has worked, the effect has been small, and what was affordable in 1990 is no longer affordable now.


See Also, "Abbott Ineffectiveness, Elementary School Edition"

See Also, "The Unsung Excellence of Dover"

See Also, "The Non-Results of Abbott Funding", a more anecdotal blog post on this same subject.  

No comments:

Post a Comment