I was intrigued a few days ago when I saw this headline in my news alerts regarding school funding.

Headline: Report: Funding helps low-performing school districts

I was particularly intrigued because the headline comes from a Connecticut newspaper where I am fully aware that the state really hasn’t done crap to substantively increase resources for low performing, or more specifically high need schools and districts.

Disclaimer: I am fully aware of this because I have been providing technical/expert assistance to local public school districts that have been persistently shortchanged by the state school finance formula (Education Cost Sharing Formula). That, and even prior to my involvement supporting these districts (and more importantly, the kids they serve) in Connecticut, I had already blogged on their plight.

So then, how can it possibly be that that a CT newspaper would print such a ridiculous headline? And where could one possibly find a “Report” that somehow validates that the state has provided funding to help low performing districts?

Well, in Connecticut, where there’s data-free drivel on education policy spewing from the headlines, there’s usually one single source for that drivel – our old friends at ConnCAN!

Yep, they’ve produced a new report! And it’s about as technically solid as many of their previous reports!

An important caveat here is that the ConnCAN report itself (the linked report) doesn’t really seem to address directly the point that is highlighted in this article – that the reforms being implemented by the Malloy administration have improved the financial conditions of districts serving high need populations.

So then where does this strange assertion come from? Did the author of the “news” (used as loosely as possible) article simply make this up – or were they fed this line by ConnCAN? I’m not sure… but the author of the article in the Middletown newspaper begins with this bold statement:

Funding made available by last year’s Public Act 12-116 has helped some of the states lowest-performing school districts, including Middletown, according to the Connecticut Coalition for Achievement Now, an education advocacy organization based in New Haven.

Then, the author of the article summarizes what are characterized as “Highlights from ConnCAN’s March 2013 Progress Report.”

I find it hard to believe the author of the article crafted these summaries on his/her own. So, let’s take these fact-challenged reformy highlights one at a time (again, on the assumption that these highlights are somehow intended to support the article’s thesis – that the reforms have somehow mitigated funding problems/disparities?):
ConnCAN Con:

School Finance: P.A. 12-116 created a Common Chart of Accounts to be implemented in 2014-15, creating across the board standards aimed at enhancing transparency in education spending. To date, the Office of Policy and Management has selected the accounting firm Blum Shapiro to develop a framework for Common Chart of Accounts development and execution.

MY REPLY

Let’s start here with simple acknowledgement that creating a common chart of accounts does little or nothing – okay, NOTHING – to enhance the equity or adequacy of educational funding across districts. So, what did the state actually do to enhance that funding? Not so much really.

Figure 1 shows the effect of the $50 million dollar increase in ECS Aid for 2012-13, when added to Net Current Expenditures (NCEP) for 2011-12. The 2011-12 NCEP distribution is shown in green dots. The changes to NCEP that would result from the additional state aid are shown in orange dots. In green dots, we see that districts like Bridgeport, New Britain, Waterbury and Meriden are significantly disadvantaged by the ECS formula in 2011-12, in terms of their resultant NCEP.

AND, perhaps more importantly, we see that “increases” to funding for 12-13 really didn’t change much!

Figure 1.

Table 1 includes NCEP for 2011-12 and the actual aid increases for 2012-13 (divided by ADM for 11-12) for Alliance Districts which include several high need districts.  I have also expressed the ECS aid increase as a percent increase over NCEP 2011-12. Most increases were less than $200 per pupil and well less than 2%.

Table 1.

Alliance District Spending & Aid Increases 12-13

ConnCAN Con:

School Choice: P.A. 12-116 increased per-pupil funding for public charter students ($10,500/FY13, $11,000/FY14, and $11,500/FY15) and allowed for the creation of 4 new state approved charters. Since then, per-pupil charter funds were cut by $300 for the FY13, and 27 letters-of-interest were submitted to the State Department of Education for launching new charters.

MY REPLY:

It is indeed true that recent adjustments to the funding formula provided more significant increases in aid to charter schools.  At best, these increases fail to alter the distribution of opportunities to Connecticut schoolchildren.  More likely, they in fact exacerbate disparities. Charters serve a relatively small share of the total student population. Most children in high need districts remain in district schools that saw negligible increase in funding. In that sense, charter funding increases have limited effect.

But, as it turns out, many of the charter schools in high need districts that received the greater increases in funding actually serve much lower need student populations (See Table 2).

Table 2. Selected Characteristics of Charter Schools in Cities where Mean % Free Lunch Exceeds 50%

Further, after removing district expenditures on transportation and special education (expenses for which host districts are primarily responsible), many charters already substantially outspent district averages (see Table 3).[1]

In short, increasing funding to charters which already outspent host districts while cream-skimming lower need students, exacerbates rather than moderating disparities in opportunity.

Table 3. Total & Comparable per Pupil Spending for Charters & Districts with Free/Reduced Lunch >50%, 2009-10, Prior to Funding Boost for Charter Schools

[1] Per Pupil Expenditures by Type: http://sdeportal.ct.gov/Cedar/WEB/ct_report/FinanceDTViewer.aspx

[2] Spending on Special Education: http://sdeportal.ct.gov/Cedar/WEB/ct_report/SpecialEducationResourcesDTViewer.aspx

[3] Percent Free or Reduced Lunch: http://sdeportal.ct.gov/Cedar/WEB/ct_report/StudentNeedDTViewer.aspx

ConnCAN Cons (lumping these last two together):

Commissioner’s Network: P.A. 12-116 gave the Commissioner of Education and the State Board of Education authority to select up to 25 of the lowest performing schools into the Commissioner’s Network school turnaround effort. Currently, 4 schools are in the Commissioner’s Network (located in Bridgeport, Hartford, New Haven, and Norwich). The state recently invited six additional schools to submit plans for inclusion in 2013-14 (located in Bridgeport, New Britain, Norwalk, Waterbury (2), and Windham).

Alliance Districts: P.A. 12-116 earmarked $39.5 million in conditional aid for the state’s 30 lowest performing school districts. So far, all 30 district plans have been approved and $39.5 million allocated.

MY REPLY

Now, in the charts above, you’ve seen the rather dramatic (cough/gag) effect that adding $50 million has on Connecticut’s high need districts through the aid formula. Well, here what we have is an even smaller amount of additional aid, to be handed out at the discretion of a single bureaucrat. Nothing systematic. Nothing substantial. Entirely discretionary, and meager.

As noted by ConnCAN, the legislation provides for 25 schools to enter the Commissioner’s network and maybe have access to some additional financial assistance.  There are far more than 25 schools in total in high need districts.  Further, each school can remain in the network for a maximum of three years, and it is unclear whether any supports would exist beyond those three years.

Let’s be absolutely clear here: Educational adequacy and equal educational opportunity a) should not be reserved for a tiny minority of schools, b) should not sunset and c) should not be at the discretion of a single political appointee.

Equally if not more likely, the various proposed structural and governance changes, coupled with new unfunded mandates, will exacerbate existing inequities across Connecticut schools and districts.   For example, many of the policy changes addressed by ConnCAN are little more than labeling schemes that merely highlight existing disparities.

Worse, the most negative and consequential labels fall disproportionately on schools in those districts already disadvantaged financially.

A substantial body of existing literature links school rating systems with local residential property values, including state accountability system assigned school grades.[2]  In short, negative labels may lead to further erosion of housing values and tax base. Further, it is likely that increased threat of state intervention and reduction of local control over schools may adversely affect local property values. The proposed reforms, lacking any substantive provision of additional resources, threaten to accelerate a downward spiral of districts already in long-run economic and educational decline.

Already, a large share of schools classified as “review” schools are not only high need schools, but high need schools concentrated in very high need, and underfunded districts (Bridgeport, Meriden, New Britain, New London & Waterbury).[5]  By contrast, the main distinction of many of the “distinction” schools identified in urban Connecticut contexts is that they serve very few of the lowest income children, few or no children with disabilities and few or no children with limited English language proficiency (See Table 4).

Meanwhile, other schools of distinction are those in the state’s most affluent suburbs.  In other words, the state has adopted a rating scheme driven primarily by student demographics to mislabel the “quality” or “effectiveness” of local public schools. Further, the rating scheme is designed to grant the state greater authority to disrupt local governance of schools, which, while the state may perceive this alternative only in positive light, local property owners and potential property owners may view it quite differently.

Table 4. Selected Characteristics of “Distinction Schools” in Cities where Mean % Free Lunch Exceeds 50%

ConnCAN Con:

Educator Evaluations: P.A. 12-116 mandated that the educator evaluation program be piloted in 8-10 sites across Connecticut. The Performance Evaluation Advisory Council (PEAC) came to an agreement that the new educator evaluation system would be implemented in all districts with flexibility in 2013-14, and the system would launch statewide with full implementation in 2014-15.

MY REPLY:

Even if one chose to accept that improved teacher evaluation systems and teacher effectiveness measures could be leveraged to better select among teachers on the labor market or in a particular district workforce, our ability to apply that leverage to improve the workforce as a whole, or achieve more equitable distribution of teaching quality would be constrained by a) the overall landscape of teacher compensation relative to other career alternatives and b)  the persistent inequities in financial resources across districts and resulting inequities teacher compensation across advantaged and disadvantaged schools and districts.

The suggestion that mandated changes to teacher evaluation alone will improve the equity and adequacy of the teacher workforce – regardless of resources – ignores that the proposed evaluation models have the potential to significantly increase job uncertainty for teachers without providing increased wages or benefits to counterbalance the risk. Increased job/career and wage expectation uncertainty, while holding wages on average, constant, is likely to lead to reduced, not increased quality of entrants to the profession.

Further, given the emerging body of evidence on the types of metrics proposed for teacher evaluation, career uncertainty is likely to be inequitably distributed, disadvantaging children in already disadvantaged districts and schools.[6]

NOTES

Recently, I responded to an absurd and downright disturbing Op-Ed by a Connecticut education reform organization that claimed that Connecticut needed to move quickly to adopt teacher evaluation/tenure reforms and expand charter schooling because a) Connecticut has a larger achievement gap and lower outcomes for low income students than Massachusetts or New Jersey and b) New Jersey and Massachusetts were somehow outpacing Connecticut in adopting new reformy policies regarding teacher evaluation. Now, the latter assertion is questionable enough to begin with, but the most questionable assertion was that any recent policy changes that may have occurred in New Jersey or Massachusetts explain why low income children in those states do better, and have done better at a faster rate than low income kids in Connecticut. Put simply, bills presently on the table, or legislation and regulations adopted and not yet phased in do not explain the gains in student outcomes of the past 20 years.

Note that I stick to comparisons among these states because income related achievement gaps are most comparable among them (that is, the characteristics of the populations that fall above and below the income thresholds for free/reduced lunch are relatively comparable among these states, but not so much to states in other regions of the country).

I’m not really providing much new information in this post, but I am elaborating on my previous point about the potential relevance of funding equity – school finance – reforms – and providing additional illustrations.

First, let’s take a look at the relationship between state and local revenues per pupil and median household income over time in Mass, CT, RI and NJ. These data are drawn from an article I published in 2010 in which I estimated a model for all states of the relationship between median household income and state and local revenue per pupil over time. In that recent article, I plotted the relationship over time for several states. Here, I have re-plotted that relationship for New Jersey, Massachusetts, Connecticut and Rhode Island.

This graph shows the slope of the statistical relationship between state and local revenues per pupil and median household income, controlling for differences in regional labor costs and economies of scale of districts. All 4 states managed to reduce or eliminate what had been positive relationships between household income and district revenues. That is, in 1990, in each state higher income districts had more revenue. By the late 1990s, Mass and NJ had reversed that pattern, and in fact had redistributed revenue such that districts with lower median household income actually had more revenue. The finance systems had become “progressive,” so-to-speak.

By contrast, the relationship between income and revenue remained more random – and on average flat – in RI and CT.  How Connecticut made so little consistent progress toward fiscal equity over time is another story for another day.

By 2009, things were pretty much the same, as the next figure shows. The next figure compares current operating expenditures per pupil and district poverty rates. Current operating expenditures include expenditure of federal funds, including Title I funds. Those federal funds tend to add marginally to the progressiveness of the system. But, what’s also important is not just whether the trendline tilts upward, but whether the pattern is systematic or predictable. In New Jersey, among districts enrolling 2,000 or more students (scale efficient), census poverty rates alone explain 47% of the variation in current spending. It’s a predictable, upward slope whereby higher poverty districts actually have and spend more per pupil. In Massachusetts, poverty explains nearly 40% of the variation in spending per pupil.

But, by contrast, in Connecticut, poverty explains only about 15% of the variation in spending across districts. Notably, a handful of high poverty districts including Bridgeport, Waterbury and New Britain have been left well behind… well below  the curve.  Further, even though Hartford and New Have have more funding, much of that funding has been channeled through a magnet school aid program, so the seemingly high position of these two districts is somewhat deceptive, and not part of any systematic effort to improve equity.

I’m doing some picking on CT here in particular, because CT has hardly provided systematic targeted support to high need districts. That said, CT is hardly the worst offender among states. Several states do far worse than CT, including PA, NY and IL in particular.  But that’s a post for another day. Those states are not sufficiently comparable in other ways to include in these comparisons. Our report on School Funding Fairness provides a full breakdown across states (www.schoolfundingfairness.org).

In the best possible case, CT has selectively targeted funding to some high need districts and has left out others. Rhode Island funding is somewhat more predictable (though deceptive because of the small number of districts) than CT but clearly less systematic than NJ or Mass.

Now, here are the longitudinal trends in student outcomes in these states, cut a few different ways and focusing on disadvantaged children. First, because I’m not a fan of comparing “low income” kids to “low income” kids even across these relatively similar economies, I take a look at children of mothers who were high school dropouts, and 8th grade math performance:

From 2000 forward (where the scores are relatively comparable over time), children of maternal high school dropouts in Mass and NJ far outpace those in CT or RI.

Next, here are the trends for children who qualify for free lunch, also on 8th grade math:

Finally, here are the trends for 4th grade reading:

In each case, low income kids in the two states that have more aggressively pursued funding equity reforms outperform and have improved their performance faster than low income kids in states that have not.

While these trends are hardly conclusive evidence [more extensive analysis is underway] of a link between the funding shifts and outcome gains, these trends are consistent with a significant body of research on the effects of school finance reforms, which I summarize in these articles/papers:

1. http://www.tcrecord.org/library/abstract.asp?ContentId=16106
2. http://www.shankerinstitute.org/images/doesmoneymatter_final.pdf

These illustrations coupled with the larger body of related research present a more compelling case for improving funding equity and adequacy than the case that has been presented thus far for other reforms – reforms which lack both any significant research base OR other compelling evidence.

Simply arguing that teacher quality varies – good teachers matter – and that we have to “fix” teacher quality says nothing of how to actually improve teacher quality or the resources required to get the job done. “Teacher quality” is not a policy.

Notably, the CT Education reform group that posted the original absurd claims later softened their arguments regarding the causes of Connecticut’s failures and Mass and NJ’s successes. But somehow, they still found it in them to pitch the same policy solutions. Yes, solutions that had nothing to do with why Mass and NJ were outperforming CT. If the original argument doesn’t hold, then neither do the proposed solutions! That’s how dumb this debate has gotten!

Let’s be absolutely clear here – During the years over which Mass and NJ saw these substantial improvements in outcomes, both states had teacher tenure systems – in both successful and less successful schools! Both states had rather traditional teacher evaluation policies in place. Both states had relatively small overall shares of children attending charter schools. Massachusetts has received some accolades for its accountability system adopted concurrent with funding reforms.

Even if the path to improved productivity and efficiency did involve reforms with any resemblance to the teacher evaluation or charter school reforms on the table, equitable and adequate financing would be a prerequisite condition for doing any of it well. Real innovation requires real investment.

Notably, the types of innovations adopted by those few charter chains which show relatively consistent records of success are rather mundane strategies which involve providing more time and intensive supplemental tutoring, typically at a substantially greater per pupil cost – much of it reflected in higher teacher salaries to support the additional time and responsibilities.   Yes, even in Connecticut those higher flying charters, while serving far less needy student populations than other schools in their host districts, are also outspending them! (and pay teachers more!)

We also have pretty good evidence from research that salaries matter (see page 7-)– that the average level of teacher salaries affects the quality of entrants to the profession – and that the relative salary paid in one district versus another may affect teacher job choice – leading to sorting of teacher credentials across districts.

By contrast we have little evidence that mass deselection of teachers on the basis of noisy and potentially biased measures of student achievement growth – without counterbalancing the risk with substantial additional reward would have any benefit whatsoever [outside of self-fulfilling simulated correlations] – and it might, in fact, do significant harm! We also have a pretty solid track record of studies suggesting few or no benefits of attaching compensation to similar performance metrics (e.g. merit pay).[1] Yet the reformy rhetoric advancing these policies through copy-and-paste template legislation persists! Is anyone taking even a brief time out to think and evaluate these proposals?

Even if we didn’t have all of this evidence for how and why money matters, wouldn’t it make sense on its face to provide schools and districts a level playing field? What possible argument is there for deciding to completely overlook substantial financial inequities in favor of policy options which, to implement properly (if that’s even possible), require equitable financial resources. Yes, real reform costs money.  Good schools cost money!

If reformy advocates really don’t believe that money has anything to do with improving schooling quality, then why do charter advocates in Connecticut (and elsewhere) push so hard for substantial increases in funding ($2,600 per pupil in CT)? Yet the underfunded high need districts stand to gain far, far less under current proposals (only about $250 per pupil). Funding those districts appropriately comes with a higher price tag than increasing charter funding because they serve a lot more students and far more needy students than CT charters. The persistent inequities faced by these districts are glaring… right down to the delivery of basic curricular options, especially in the neediest, least well resourced districts. Larger class sizes… Advanced and enriched curricular options… Teacher Salaries… Concentration of novice teachers.

These are the real differences between well resourced, high performing and poorly resourced, low performing districts in Connecticut.  Both groups of districts pay teachers based on degrees and experience. Both sets of districts have tenure systems. I suspect that both sets of districts do little systematically different regarding teacher evaluation (which might be improved, but not by some ill-conceived state legislation).  So clearly none of that stuff has much of anything to do with which districts are succeeding and which are not ! The districts are substantively  different in terms of who they serve, and in terms of the resources (by the measures noted above) they have to serve them!

I’ve long argued that it seems rather hypocritical to hear staunch charter advocates argue that more money wouldn’t help traditional public schools, and then those same charter advocates set out to help their favored charters substantially outspend the nearby public schools (meanwhile, invariably serving less needy, less costly children to educate).

If money has nothing to do with improving schooling quality – or providing high quality schooling – then why is the average tuition of private independent day schools in Connecticut typically well above (around $25k elem, up to $35k HS) current spending levels of nearby public districts (where tuition covers only a portion of current spending).[2] Money – and what it buys – clearly means something to the consumers of what I might call luxury schooling.  Nationally, private independent schools spend on average 1.96 times average spending of public districts in their same labor market. [3] Meanwhile, these schools DO NOT generally fire teachers at will. They do typically pay based largely on seniority and degree level. What do they do? Well, for one thing, they provide small class sizes [whether that’s the most efficient allocation or not is a separate question, but that is what they choose to do… and what they often advertise]!

It is about the money. Even/especially those disingenuous arguments that it’s NOT about money? Those arguments perhaps, are the ones that most clearly indicate that it IS about money – and a complete unwillingness to acknowledge the importance of money to anyone else but one’s self. Strangely, we’ve reached a point in the discourse where it’s all about finding any rationale we can to NOT give money to the schools and districts that need it most while continuing to blame them, their teachers and the children they serve for the persistent cycle of poor performance – and where we define their performance as poor by comparison to their more advantaged and better resourced peers! Isn’t anyone else seeing the absurdity?[4] [starve them… shift money away from them… then blame them for doing even worse?]

Clearly the media isn’t catching it. I’m less frustrated by the absurdity of these arguments (I rather expect it in the political sphere) than the fact that the media is either complicit in advancing the silliness, or intellectually incapable of seeing through it.

Those who continue to ignore outright the role of money in schooling in favor of reformy window dressing – meanwhile leveraging every opportunity to access a greater share of the federal, state and local tax dollar, present a laughable case for their preferred reforms.  But for some reason, I’m finding it really hard to laugh.

=======

[1]Glazerman, S., Seifullah, A. (2010) An Evaluation of the Teacher Advancement Program in Chicago: Year Two Impact Report. Mathematica Policy Research Institute. 6319-520
Springer, M.G., Ballou, D., Hamilton, L., Le, V., Lockwood, J.R., McCaffrey, D., Pepper, M., and Stecher, B. (2010). Teacher Pay for Performance: Experimental Evidence from the Project on Incentives in Teaching. Nashville, TN: National Center on Performance Incentives at Vanderbilt University.
Marsh, J. A., Springer, M. G., McCaffrey, D. F., Yuan, K., Epstein, S., Koppich, J., Kalra, N., DiMartino, C., & Peng, A. (2011). A Big Apple for Educators: New York City’s Experiment with Schoolwide Performance Bonuses. Final Evaluation Report. RAND Corporation & Vanderbilt University.  http://www.americanprogress.org/issues/2011/04/pdf/class_size.pdf

[2] http://www.brunswickschool.org/admissions/fees-financial-aid/, http://www.chasecollegiate.org/page.cfm?p=208, http://www.fairfieldprep.org/page.cfm?p=24, http://www.greenwichacademy.org/podium/default.aspx?t=32259, http://www.klht.org/podium/default.aspx?t=2170, http://www.stanwichschool.org/admissions/tuition.asp, http://woosterschool.org/admissions/tuition-fees

[3] Baker, B. (2009). Private schooling in the U.S.: Expenditures, supply, and policy implications.
Boulder and Tempe: Education and the Public Interest Center & Education Policy Research
Unit. Retrieved [date] from http://epicpolicy.org/publication/private-schooling-US

[4] An argument taken to its extremes in a recent NJDOE report: https://schoolfinance101.wordpress.com/2012/02/24/how-not-to-fix-the-new-jersey-achievement-gap/

A few days ago, I responded to an utterly silly CT Ed Reform op-ed which argued that poverty doesn’t really matter so much, nor does funding (by omission), and that Massachusetts and New Jersey do better than Connecticut on behalf low income kids because they’ve adopted accountability and teacher evaluation reforms in the past few years. Thus, the answer is for Connecticut to follow suit by adopting SB 24 in its original form. To be clear, NJ has absolutely not adopted anything like SB 24.  Here’s a key section of that op-ed:

We think folks would be hard-pressed to argue that low-income students right over the border in Massachusetts or New Jersey face very different circumstances at home than the low-income students in Connecticut.  So, what actions have our neighboring states taken to address their achievement gaps that Connecticut hasn’t?  Put bluntly, they have adopted education reform policies very similar to the ones proposed in Governor Malloy’s original education reform bill.  They have adopted or implemented policies that evaluate teachers on the basis of student performance, that rank schools and districts within a tiered intervention framework, and that provide the Commissioner with the authority to intervene in the lowest performing schools and districts.

In my previous post, I already pointed out a few simple realities… like the fact that both Massachusetts and New Jersey have systematically tackled school funding equity over time, where Connecticut has not. I also suggested that it might be rather foolish to argue that policies considered and/or adopted but not even really implemented yet were the cause of improvements to New Jersey and Massachusetts low income student progress.

So, here’s just one more graph to drive home that point:

http://nces.ed.gov/nationsreportcard/naepdata/dataset.aspx

Yes – Mass and NJ do better than CT, across all student groups, including lower income students. And yes, Mass and NJ have shown higher rates of growth in student outcomes, across all student groups. And you know what, much if not most of that growth has occurred prior to considering, piloting and partially implementing new policies.

So… to put it really simply… it’s pretty darn unlikely that Mass and NJ do better than CT because of recent policy developments.

Further, across all three states, poverty continues to matter.

Children who do not qualify for either free or reduced lunch outperform those who do.

Children who qualify for reduced lunch (185% poverty income level) outperform those who qualify for free lunch (130% poverty income level).

Notably, children qualifying for free lunch in Massachusetts have surpassed those qualified for reduced lunch in CT and those qualified for free lunch in NJ are catching up with those qualified for reduced lunch in CT.

In fact, gaps in the other two states remain relatively large as well because of the growth in outcomes of higher income students which has largely paralleled the growth among lower income students.

This utterly absurd post appeared yesterday on the CT Ed Reform blog:

http://ctedreform.org/blog/2012/04/poverty-is-not-to-blame-ct%E2%80%99s-low-income-students-rank-48th-in-the-nation-while-ma%E2%80%99s-rank-2nd/

Essentially, the argument goes:

1. CT’s achievement gap is worse than achievement gaps in states like Massachusetts and New Jersey and in particular, CT’s low income students perform less well than low income students in those states.
2. Massachusetts has recently adopted reforms to teacher evaluation, which is obviously why Massachusetts has a smaller achievement gap and better performance among low income students. (the post cites New Jersey as well)
3. THEREFORE, WE KNOW THAT POVERTY IS NOT THE ISSUE…. IT’S TEACHER EVALUATION (and Charter schools, and other reformy stuff)!
4. Therefore, the solution is to pass SB24 in its original form, which includes such fun things as student test based evaluation of teachers (3 good years to tenure, 2 bad and you’re out), additional funding for and expansion of charter schools – which we know can overcome this pesky poverty distraction!

Now, before I even begin here, I found it most absurd that this particular Ed Reform posting attributed progress made in Massachusetts over the past two decades, to policies adopted in the past two years! As they put it:

We think folks would be hard-pressed to argue that low-income students right over the border in Massachusetts or New Jersey face very different circumstances at home than the low-income students in Connecticut.  So, what actions have our neighboring states taken to address their achievement gaps that Connecticut hasn’t?  Put bluntly, they have adopted education reform policies very similar to the ones proposed in Governor Malloy’s original education reform bill.  They have adopted or implemented policies that evaluate teachers on the basis of student performance, that rank schools and districts within a tiered intervention framework, and that provide the Commissioner with the authority to intervene in the lowest performing schools and districts.

That’s just funny! Ridiculous in fact, making me wonder if this really was just an April fool’s joke.   Let’s make this really plain and simple.

Policies adopted in Massachusetts in the past two years and not yet even fully implemented did not cause low income children in Massachusetts to outperform low income children in Connecticut between 1992 and 2009!

And about New Jersey, I’m not quite sure what policies (legislation) they are talking about, since legislation regarding teacher evaluation is still in its early incubation stages. Pilot studies have begun in a handful of districts but  to suggest that pilot studies being implemented and evaluated now somehow explain past performance of low income students in New Jersey is, well, just dumb.

Now, Massachusetts and New Jersey have in fact implemented reforms that Connecticut has not – SCHOOL FINANCE REFORMS (in Mass, coupled with accountability reforms in the 1990s). Both states have more systematically targeted funding to higher need districts. And for a review of the literature on the effects of such school finance reforms, with specific references to New Jersey and Massachusetts, see:

1. http://www.tcrecord.org/library/abstract.asp?ContentId=16106
2. http://www.shankerinstitute.org/images/doesmoneymatter_final.pdf

While Connecticut has selectively driven magnet aid to Hartford and New Haven, Connecticut has left other high need districts out entirely. Further, in recent years, as shown in a previous post, Connecticut charter schools have substantively segregated students by income within Hartford, New Haven and Bridgeport.

Here’s a quick snapshot, using 2009 data, of the relationship between current spending per pupil and U.S. Census poverty rates for districts enrolling over 2,000 pupil within New Jersey, Massachusetts and Connecticut. The notable feature of these graphs, indicated by the r-squared value is that in both Massachusetts and New Jersey, current spending per pupil is far more predictably a function of differences in local district poverty rates (adjusted for regional cost variation). In New Jersey and Massachusetts, poverty variation explains more than a third, to nearly half the variation in per pupil spending, and in Connecticut, less than 1/6!

Financial data: http://www.census.gov/govs/school/

Poverty data: http://www.census.gov/did/www/saipe/data/schools/data/index.html

Now, on to other issues: That achievement gap!

Yes, Connecticut does have a relatively large achievement gap, but that gap has to be put in context with similar states, as I explain here. The short version of this story is that the low income- non-low income achievement gaps across states are largely a function of the income gaps between the two groups. Here’s my graph of that relationship:

In the upper right hand corner of this graph are the states with both large income gaps between poor and non-poor kids and with large achievement gaps between them. Yes, Connecticut’s gaps are larger than those of Mass or New Jersey, and those are perhaps the most relevant comparison states (the post got that right – but that’s about all).

We know the correlates of student achievement across CT schools: Poverty!

As it turns out, across these three states, in each case, lower income students perform less well on NAEP. Children qualified for reduced lunch perform less well than those not qualified for subsidized meals at all, and children qualified for free lunch perform less well than those who qualify for reduced price lunch. That’s why, in some cases, I choose t parse these populations in comparisons where most schools serve children below the upper threshold.

Data source: http://nces.ed.gov/nationsreportcard/naepdata/dataset.aspx

Data source: http://nces.ed.gov/nationsreportcard/naepdata/dataset.aspx

It also turns out that if we just go nutty with lots of different measures across Connecticut districts to identify those factors that are most highly correlated with student achievement measures (all data can be found here: http://sdeportal.ct.gov/Cedar/WEB/ct_report/DTHome.aspx & http://www.nces.ed.gov/ccd/bat) we find that various measures of poverty or household income are pretty darn highly associated with student outcome measures! This would seem to suggest that perhaps poverty and measures related to it do likely matter in some way. We’re talking about correlations near and above .80 here between % free/reduced and CMT scores across districts!

We know which districts have greater needs, by various measures!

The reality is that by any number of measures of income or poverty, we know which Connecticut districts have greater student needs, low income and in many cases more children with Limited English Language proficiency.  Here are the scatterplots of the relationships between poverty, income and ELL measures used in the above correlation analysis.

Yep. It’s all pretty straightforward. Various measures of income and poverty are pretty highly related across Connecticut. It’s a highly socioeconomically and racially segregated state. And student outcome measures remain highly correlated with socio-economic measures and with racial composition of school districts!

Malloy’s plan does little or nothing to help them financially

We also know from my previous posts that the Malloy plan does little or nothing to infuse additional resources into the highest need districts. Here it is again!

Additional data sources suggest that CT charters serve fewer needy kids and spend more per pupil than surrounding district schools

But, the Malloy plan does include substantial boost in funding for charter schools which I have shown in previous posts, tend to serve the less needy kids within the highest need settings in the state.  Further, I have shown in those earlier posts that Connecticut charter schools don’t appear to be systematically financially disadvantaged when compared to traditional public schools.

I recalled the other day that the new U.S. Department of Education school site data set released a short while ago includes per pupil spending figures for charter schools in some states. Among those states is Connecticut. The report and data also include information on shares of children by school who are low income.

Here are a few quick snapshots of how Connecticut charter schools in Hartford, New Haven and Bridgeport compare in terms of spending and poverty to “regular” public schools in each of those host districts.

First, without charter names:

Data source: http://www2.ed.gov/about/offices/list/opepd/ppss/reports.html#comparability-state-local-expenditures

[Note that the data set has the variable labeled as “school poverty rate” when in fact it is, I believe, a % free or reduced lunch measure. I’ve left the data label as it is in the original data set]

Now, with the names:

Data source: http://www2.ed.gov/about/offices/list/opepd/ppss/reports.html#comparability-state-local-expenditures

[Note that the data set has the variable labeled as “school poverty rate” when in fact it is, I believe, a % free or reduced lunch measure. I’ve left the data label as it is in the original data set]

A few things are notable here.

First, overall, there’s simply no upward tilt in per pupil spending by school poverty rate. That said, I’m only looking here within higher poverty settings (with the trendline determined by the regular public schools only).  In other words, higher poverty schools don’t generally have more resources per pupil than lower poverty ones within these cities, but my experience with similar data in other settings indicates that these variations are most often explained by the distribution of children with disabilities (also scarce in CT charters).

Second, as I illustrated in my previous post, the charter schools in these cities stand out in terms of the populations they serve (by low income status). New Haven schools have somewhat of a spread of low income rates, but Hartford and Bridgeport are all crunched up against the 100% mark (not all at 100% though, and with some more spread when I use free lunch only). In each case, charter % low income is lower than most regular schools in the host district.

Third, consistent with my previous analysis (but likely because the state reported the data to USDOE), many (more than not & all Achievement 1st schools) charter schools appear to be spending not only more than schools serving similar student populations (by income status), of which there are very few in these settings, but also more than district schools serving much lower income student populations.

These data certainly  raise questions about the validity of the current policy push in Connecticut and raise even more questions about the stated reasons for that push. That is, to the extent that anyone truly believes the absurd rhetoric that test-based teacher evaluation policies and expanding higher spending lower poverty charter schools are the solution to Connecticut’s achievement gap.

Connecticut’s SB 24 appears to be little more than boilerplate reformy legislation which, like similar legislation in other states, creates a massive smokescreen concealing the very real problems facing Connecticut school districts. I addressed in a previous post my concern that SB24’s emphasis on charter expansion as a solution for high poverty districts is misguided, mainly because most of those successful charter schools in CT are currently achieving their successes at least in part by NOT serving high poverty populations. And another part may be the additional resources of these schools, used for such things as increased school time, supported by increased teacher salaries.  But SB24 comes with few resources attached. The other major elements of SB24 involve teacher “effectiveness” with significant emphasis on use of student performance measures for teacher evaluation. For numerous posts on this topic, see: https://schoolfinance101.wordpress.com/category/race-to-the-top/value-added-teacher-evaluation/

A few points are in order before I move on.

First, even if we make value added measures about 20% of an evaluation system, and observations and other measures cover the rest, if the value-added measures vary most (which they are likely to, simply because they will be reported as statistically norm-referenced), then the value added measures likely become the tipping point more often than not. This is hugely problematic, given our inability to fully remove bias from these measures, and the fact that they remain so damn noisy as to hardly be useful at all (and are easily manipulated to yield different results for individual teachers)

Second, arguing that somehow using these noisy, potentially biased measures for personnel management or even mass deselection of teachers will somehow improve the equity of the distribution of teachers across advantaged and disadvantaged schools is simply absurd! This is especially the case if absolutely no attention is paid to existing underlying disparities in working conditions and teacher compensation.

The Real Connecticut Problem(s)

So, let’s take a look at what’s really going on in Connecticut regarding the distribution and compensation of teachers. But, let me begin with a bit of background literature on the relationship between funding and teacher quality. Rather than reinvent the wheel here, allow me to rely on a section of a policy brief I wrote last fall for Shanker Institute:

The Coleman report looked at a variety of specific schooling resource measures, most notably teacher characteristics, finding positive relationships between these traits and student outcomes. A multitude of studies on the relationship between teacher characteristics and student outcomes have followed, producing mixed messages as to which matter most and by how much.[i] Inconsistent  findings on the relationship between teacher “effectiveness” and how teachers get paid – by experience and education – added fuel to “money doesn’t matter” fire. Since a large proportion of school spending necessarily goes to teacher compensation, and (according to this argument) since we’re not paying teachers in a manner that reflects or incentivizes their productivity, then spending more money won’t help.[ii] In other words, the assertion is that money spent on the current system doesn’t matter, but it could if the system was to change.

Of course, in a sense, this is an argument that money does matter. But it also misses the important point about the role of experience and education in determining teachers’ salaries, and what that means for student outcomes.

While teacher salary schedules may determine pay differentials across teachers within districts, the simple fact is that where one teaches is also very important in determining how much he or she makes.[iii] Arguing over attributes that drive the raises in salary schedules also ignores the bigger question of whether paying teachers more in general might improve the quality of the workforce and, ultimately, student outcomes. Teacher pay is increasingly uncompetitive with that offered by other professions, and  the “penalty” teachers pay increases the longer they stay on the job.[iv]

A substantial body of literature has accumulated to validate the conclusion that both teachers’ overall wages and relative wages affect the quality of those who choose to enter the teaching profession, and whether they stay once they get in. For example, Murnane and Olson (1989) found that salaries affect the decision to enter teaching and the duration of the teaching career,[v] while Figlio (1997, 2002) and Ferguson (1991) concluded that higher salaries are associated with more qualified teachers.[vi] In addition, more recent studies have tackled the specific issues of relative pay noted above. Loeb and Page showed that:

“Once we adjust for labor market factors, we estimate that raising teacher wages by 10 percent reduces high school dropout rates by 3 percent to 4 percent. Our findings suggest that previous studies have failed to produce robust estimates because they lack adequate controls for non-wage aspects of teaching and market differences in alternative occupational opportunities.”[vii]

In short, while salaries are not the only factor involved, they do affect the quality of the teaching workforce, which in turn affects student outcomes.

Research on the flip side of this issue – evaluating spending constraints or reductions – reveals the potential harm to teaching quality that flows from leveling down or reducing spending. For example, David Figlio and Kim Rueben (2001) note that, “Using data from the National Center for Education Statistics we find that tax limits systematically reduce the average quality of education majors, as well as new public school teachers in states that have passed these limits.”[viii]

Salaries also play a potentially important role in improving the equity of student outcomes. While several studies show that higher salaries relative to labor market norms can draw higher quality candidates into teaching, the evidence also indicates that relative teacher salaries across schools and districts may influence the distribution of teaching quality. For example, Ondrich, Pas and Yinger (2008) “find that teachers in districts with higher salaries relative to non-teaching salaries in the same county are less likely to leave teaching and that a teacher is less likely to change districts when he or she teaches in a district near the top of the teacher salary distribution in that county.”[ix]

With regard to teacher quality and school racial composition, Hanushek, Kain, and Rivkin (2004) note: “A school with 10 percent more black students would require about 10 percent higher salaries in order to neutralize the increased probability of leaving.”[x] Others, however, point to the limited capacity of salary differentials to counteract attrition by compensating for working conditions.[xi]

Finally, it bears noting that those who criticize the use of experience and education in determining teachers’ salaries must of course produce a better alternative, and there is even less evidence behind increasingly popular ways to do so than there is to support the policies they intend to replace. In a perfect world, we could tie teacher pay directly to productivity, but contemporary efforts to do so, including performance bonuses based on student test results,[xii] have thus far failed to produce concrete results in the U.S. More promising efforts to measure productivity, such as new teacher evaluations that incorporate heavily-weighted teacher productivity measures based on their students’ test scores, are still a work in progress, and there is not yet evidence that they will be any more effective (or cost-effective) in attracting, developing or retaining high-quality teachers.

To summarize, despite all the uproar about paying teachers based on experience and education, and its misinterpretations in the context of the “Does money matter?” debate, this line of argument misses the point. To whatever degree teacher pay matters in attracting good people into the profession and keeping them around, it’s less about how they are paid than how much. Furthermore, the average salaries of the teaching profession, with respect to other labor market opportunities, can substantively affect the quality of entrants to the teaching profession, applicants to preparation programs, and student outcomes. Diminishing resources for schools can constrain salaries and reduce the quality of the labor supply. Further, salary differentials between schools and districts might help to recruit or retain teachers in high need settings. In other words, resources used for teacher quality matter.

So then, how does this all play out in Connecticut? By the reformy rhetoric being so casually tossed about one would think that all of those urban CT teachers must already be being paid lavishly and certainly more than enough than would be required to get the best and brightest CT college grads to want to teach in Bridgeport or New Britain.

Let’s start with wages relative to non-teaching professions. Allegretto, Corcoran and Mishel identify Connecticut as having a relatively average teaching penalty.

In Connecticut, the average weekly wages of teachers are about 77.6% of the average weekly wages of similarly educated non-teachers.

But, we also know that Connecticut’s good schools and districts are pretty darn good. So perhaps the issue isn’t so much about the average, but about the disparities. Besides, the core rhetoric around the proposed reforms seems to be much about the achievement gap in CT, which is indeed large even when corrected for the income gap.

Let’s quickly revisit my representation of which districts in CT are most disadvantaged, when we look at the relationship between cost and need adjusted per pupil expenditures and average current outcomes:

Expressing the funding as difference from the average, and throwing some cutpoints into the picture to create some fun groups for comparison, I get:

So, I’ve got the advantaged districts which have high adjusted spending and high outcomes. I’ve got my overall group of disadvantaged districts which have low adjusted spending and low outcomes, and a particularly screwed subset of these districts which I call severely disadvantaged (including Bridgeport and New Britain).

Now, recall, that the funding side of the Malloy plan isn’t going to do a whole lot to help out these districts.

Notice that huge infusion of funding represented by the red triangles relative to prior year Net Current Expenditures. Oh wait. There really isn’t any. But what’s funding got to do with it anyway? (go back and read the above section!)

So, I’ve taken the teacher level salary and characteristics data to estimate two different models of the differences in teacher characteristics between the advantaged and disadvantaged districts and between the advantaged and severely disadvantaged districts. First, let’s look simply at salary parity at constant teacher characteristics. NOTE THAT IT WOULD TAKE NOT ONLY AN EQUAL, BUT HIGHER SALARY, FOR EXAMPLE, TO GET TEACHERS WITH SPECIFIC QUALIFICATIONS TO WORK IN BRIDGEPORT AS OPPOSED TO WESTPORT.

Here’s the salary model, with comments in the margins:

All else equal, teachers in disadvantaged districts are still behind their peers in advantaged districts within the same labor market. Reformy platitudes and test based evaluation will not fix that!

Now, here’s a logistic regression of the likelihood that a teacher is a novice, or in his/her first 3 years of teaching. This is a commonly used marker for teacher quality inequity, because a substantial body of literature has found that concentrations of novice teachers (i.e. teachers with less than 3 or 4 years of experience) can have significant negative effects on student outcomes.[1] Rivkin, Hanushek, and Kain (2005) find that teacher experience is important in the first two years of a teaching career (but not thereafter).[2]  Hanushek and Rivkin note that: “we find that identifiable school factors – the rate of student turnover, the proportion of teachers with little or no experience, and student racial composition – explain much of the growth in the achievement gap between grades 3 and 8 in Texas schools.”[3] Notably, evidence from a variety of state and local contexts, provides a consistent picture that higher concentrations of novice teachers are associated with negative effects on student outcomes.

Here are the models:

So, kids in classrooms in severely disadvantaged or generally disadvantaged districts are each about 20% more likely to face novice teachers. Note that they are also more likely to be in larger classes, specifically if they are in the severe disparity group!

Again, SB24’s reformy platitudes will do nothing to remedy this disparity.

Put simply, the SB24 teacher effectiveness provisions are a massive smokescreen that do little or nothing to address persistent underlying disparities across CT districts. Worse, the misguided emphasis on reducing job security and focusing on problematic performance metrics will likely do more harm than good for children in the most disadvantaged districts.