Lederman v. King

OPINION OF THE COURT Roger D. McDonough, J. Petitioner seeks a judgment: (1) setting aside or vacating the Growth Score and Rating (growth score) of petitioner of 1 out of 20, and the identification of petitioner as “Ineffective” for school year 2013-2014; and (2) declaring that the New York State Growth Measures (growth measures) are arbitrary and capricious and an abuse of discretion, and permanently enjoining the use of said “growth measures” unless they are modified to rationally evaluate teacher performance. Respondents oppose the petition in its entirety. — 888 — Background Petitioner is a fourth grade teacher employed by the Great Neck Public School District in the State of New York. She has been so employed since September of 1997. For the school year 2013-2014, she received a growth score of 1 out of a possible 20 points. Said score correlates to a rating of “Ineffective.” A score/ rating sheet defines “Ineffective” as: “Results are well-below State average for similar students.” For the school year 2012-2013, she received a growth score of 14 out of 20 points. Said score correlates to a rating of “Effective,” defined as: “Results meet State average for similar students.” Petitioner, via her counsel, confirmed with the New York State Education Department that neither an administrative appeal nor an appeal to the Education Commissioner was available to challenge her growth score. The instant proceeding ensued. Procedural Background Respondents moved to dismiss the petition for lack of standing. Petitioner cross-moved for permission to conduct discovery and to supplement her petition. The court denied the motion to dismiss and reserved decision on the discovery issue. Additionally, the court granted petitioner permission to supplement her petition. A briefing schedule was established, respondents served their answer and the court heard oral argument on the petition. Additionally, the parties reached a stipulation as to certain issues of confidentiality. Prior to the petition’s final return date, petitioner brought a motion (unsealing motion) to unseal certain data as well as for leave to potentially file an amended petition. Respondents opposed all of the relief requested in the unsealing motion. During the pendency of the court’s consideration of the petition and the unsealing motion, petitioner attempted to make additional submissions a part of the record. The court requested, on notice to petitioner, respondents’ input as to whether the additional submissions should be considered by the court. Respondents objected to the untimely submissions and argued that the court, if it were to permit and consider any new submissions, should also consider the December 2015 actions of the Board of Regents. The respondents contended that said actions rendered the petition, entirely or in portions, moot and also signaled respondents’ willingness to potentially compromise the remaining claims at issue. Accordingly, respon — 889 — dents asked the court to delay issuance of any decision until either a formal briefing schedule was set for the mootness issue or potential settlement was explored. Over petitioner’s objection, the court agreed to delay the decision to allow further briefing as well as settlement talks. The settlement talks proved unsuccessful and final submissions have been received as to the potential mootness issue. Additionally, the court indicated that it would consider petitioner’s post-return date submissions from November and December of 2015. Discussion Mootness As set forth above, petitioner’s relief requests are twofold. Specifically, she seeks a judgment: (1) setting aside or vacating the growth score and rating of 1 out of 20, and the identification of petitioner as “Ineffective” for school year 2013-2014; and (2) declaring that the New York State Growth Measures are arbitrary and capricious and an abuse of discretion, and permanently enjoining the use of said “growth measures” unless they are modified to rationally evaluate teacher performance. As to the first category of requested relief, respondents maintain that the petition is now moot based on: (1) petitioner’s 2014-2015 growth score and rating;2 (2) the lack of public access to petitioner’s 2013-2014 growth score and rating; and (3) the absence of any adverse employment actions, loss of merit pay, etc., by petitioner. In opposition, petitioner principally points out that the specific relief she sought was a setting aside/ vacating of her growth score and rating. As said relief has not arisen from the Regents’ actions, petitioner maintains that her first category of requested relief has not been rendered moot. The court finds that the petition is not moot as to the first category of requested relief. Petitioner has not received all the relief she is arguably entitled to in this CPLR article 78 proceeding (see Matter of Ponder v Annucci, 128 AD3d 1255 [3d Dept 2015]). Additionally, respondents’ arguments were largely focused on petitioner’s standing in this matter. The court finds no legal basis to revisit its previous standing determination. As to the second area of requested relief, respondents maintain that petitioner’s challenges to the growth model have — 890 — been rendered moot by the recent regulatory reforms. In support, respondents rely upon: (1) the implementation of an appeal procedure for an educator to challenge his/her growth score for the 2014-2015 school year and thereafter; (2) the implementation of a four year moratorium on the use of certain state assessments and the state-provided growth model in annual professional performance reviews through the 2018-2019 school year;3 and (3) the implementation of a new state-provided growth model after the 2018-2019 school year. Respondents also note that the Regents have explicitly stated that the growth model that the petitioner is challenging will not be used as it currently exists. As such, respondents argue that the second area of requested relief has been rendered moot. In opposition, petitioner relies upon: (1) the continued use of the challenged scores; (2) the fact that the regulations are subject to modification at any time; (3) the possibility that at the expiration of the regulations the original growth model will be used again; and (4) the fact that scores can still be provided to parents and collected as aggregate data. Alternatively, petitioner argues that she’s met all components of the exception to the mootness doctrine. The court finds that the petition is moot as to the second category of requested relief. With new regulations already in place, the court’s declaration as to the arbitrariness and capriciousness of the supplanted regulations would have no practical, binding effect on the parties, the Regents, the legislature, etc. (see Saratoga County Chamber of Commerce v Pataki, 100 NY2d 801, 811 [2003]). Additionally, the court has not been persuaded that the exception to the mootness doctrine applies. Said conclusion is based on, inter alia, the fact that the new regulations are unlikely to evade review (see id.). The court cannot envision how the new regulations could evade judicial challenge from a teacher with proper standing. Additionally, the court and the aggrieved teacher will likely have the added benefit of a full administrative appeal and determination. Based on all of the foregoing, the court finds that the second category of requested relief has been rendered moot by — 891 — the actions of the Regents. As such, the court declines to render a declaratory judgment as to the subjects at issue in the second category of relief. Petitioner’s 2013-2014 Growth Score and Rating The overall gist of petitioner’s argument is that respondents’ system unjustly punishes teachers whose students consistently score substantially in excess of the state’s standards. Petitioner points out that, for the 2013-2014 school year: (1) 61.1% of her students met or exceeded the English language arts standards; and (2) 72.2% of her students met or exceeded the math standards. Petitioner compares these figures to the New York State average of 31% and is stupefied as to how she could have received a growth score of 1 out of 20. Her incredulity increases when she notes that for the 2012-2013 school year, her students’ scores were 68.75% for both English language arts standards and math. Petitioner argues that the slight improvement in math and the slight decline in English between the years could not possibly correlate with a drop in her growth score from 14/20 to 1/20 and a drop in her rating from effective to “Ineffective.” In sum, petitioner principally argues that her students have objectively performed far in excess of state standards for both school years 2012-2013 and 2013-2014. Petitioner has provided the court with an affidavit from her school’s principal. Therein, the principal asserts that the English language arts and math scores of petitioner’s students constitutes “extraordinarily effective teaching.” The petitioner also provided the court with an affidavit from the superintendent of petitioner’s school district. The superintendent echoes the words of the principal and argues that the scores at issue constitute “extraordinarily effective teaching.” In a subsequent submission, petitioner argues that the computer model used by respondents is biased against teachers with high performing students. She also points to the wide, inexplicable variation in her scores from 2012-2013 to 2013-2014 as proof that the computer model utilized by respondent is unreliable. Petitioner provided the court with an affidavit from professor Linda Darling-Hammond of Stanford University. Therein, the professor concludes that petitioner’s growth score and rating for 2012-2013 are utterly irrational. She noted, inter alia, that studies of value added models (VAMs) indicate that teachers account for only about 1-14% of the variability in test scores. The professor also points out that New York’s VAM does not at — 892 — tempt to measure control for school, classroom or home variables. She also notes that even the controls for variables such as poverty, disability or English language learners are not measured in a particularly nuanced fashion. The professor also discusses the difficulty VAM systems have in measuring both high-achieving and low-achieving students because the state tests cannot measure growth for students who are already above grade level in their skills or who fall considerably below. She notes that petitioner’s classroom included a substantial proportion of high-achieving students. The professor also notes that petitioner’s classroom sizes were very small and that smaller sample sizes produce larger error ranges. In sum, she maintains that the large swing from 14/20 to 1/20 indicates that New York’s VAM is not accurately measuring teaching effectiveness.4 Petitioner also supplied the court with an affidavit from professor Aaron M. Pallas, Ph.D., of Columbia University. Professor Pallas indicates that, under New York’s system, the percentages of teachers falling into four different categories of growth scores is predetermined. Seven percent will be in the top category (highly effective), 77% will be in the second category (effective), 9% in the third category (developing) and 7% in the fourth category (ineffective). He further points out that these four categories will remain consistent from year to year, regardless of whether the performance of students dramatically rose or dramatically fell from the previous year. The professor concludes that said model is irrational and unreasonable. Petitioner also supplied the court with an affidavit from professor Audrey Amrein-Beardsley of Arizona State University. The professor concludes that petitioner’s 2013-2014 growth score and rating are utterly irrational. She initially notes that the VAM used by New York has never been externally, peer-reviewed. The professor also points out that students in the well-above average and/or gifted ranges of performance are unable to demonstrate growth akin to lower performing students. She also points out that the extreme drop in petitioner’s growth score from 2012-2013 to 2013-2014, while the scores of her students are statistically similar, is strong evidence of the lack — 893 — of reliability of the testing and model being used by respondents. Petitioner has provided the court with an affidavit from Dr. Carol Burris, a principal from South Side High School in Rock-ville Centre, New York. Dr. Burris notes that about a third of teachers saw their growth scores change by an entire rating category between 2011-2012, 2012-2013 and 2013-2014. Additionally, she appears to speculate that the State Education Department has given extra “points” to teachers of disadvantaged students regardless of the student’s prior level of achievement. Dr. Burris cites this example and certain figures for New York City teacher growth scores as evidence of a potential bias in the system against suburban teachers. Petitioner has also supplied the court with an affidavit from Brad Lindell, Ph.D., a clinical/school psychologist and a statistical and research consultant. Dr. Lindell concludes that petitioner’s 2013-2014 growth score and rating are irrational. He also notes that even if 100% of New York’s students were proficient on the state tests and had shown significant growth from a prior year, a predetermined portion of teachers in the state would still be deemed ineffective. He argues that this predetermined result is simply not rational. The petitioner has also supplied the court with eight affidavits from parents of students and two affidavits from former students attesting to her excellence as a teacher and the surprise the affiants experienced at petitioner’s growth score of 1 out of a possible 20. In response, respondents have provided the court with an affidavit from Daniel Sherman, a managing economist employed by the entity that developed New York’s growth model. He has provided the court with information indicating that for the 2013-2014 school year, 26 out of 32 of petitioner’s students performed “less well than other similar students in the State.” The remaining six performed as well or better than their expected scores. Mr. Sherman notes that other teachers with similar students did receive higher ratings and points on the growth sub-component than petitioner because their students showed more academic growth on state assessments. He selected 100 teachers in the state with the students most similar (in terms of mean prior scores in math and English language arts) and found that five were rated highly effective, 80 were rated effective, 12 were rated developing and three were rated ineffective. In sum, Sherman concludes that New — 894 — York’s growth model is a rational approach to measuring student growth. Respondents also provided the court with an affidavit from Ira Schwartz, an Assistant Commissioner at the New York State Education Department. Mr. Schwartz contends that petitioner is confusing student growth with student achievements. In this vein, he maintains that petitioner’s reliance on her student’s proficiency rates to statewide proficiency rates are not relevant. Rather, the focus on the student growth issue is a comparison of a student’s growth compared to other similar students. He stresses that the model in place appropriately measures the change in student performance between two points in time. Mr. Schwartz also stresses that petitioner’s students have not grown as much as predicted in comparison to similar students across New York State. In sum, Schwartz concludes that New York’s growth model is rational, reasonable and well-supported. In reply, petitioner maintains that respondents’ submissions have established that: (1) the process by which her growth score was assigned is not transparent; and (2) the process by which points were calculated for her growth score was not available to her at the beginning of the 2013-2014 school year. Petitioner also describes the submissions as completely unresponsive to the litany of expert affidavits submitted on her behalf. In particular, she notes the absence of any meaningful challenge to her experts’ conclusions regarding radical score swings from year to year and the unreliable nature of VAMs from a statistical point of view. The petitioner also challenges the credentials of Mr. Schwartz and Mr. Sherman and faults respondents for failing to challenge the numerous affidavits attesting to petitioner’s excellence as an educator. Finally, petitioner again stresses the irrationality of a system that predetermines that 7% of teachers are ineffective regardless of statewide improvements or declines in student scores. Petitioner also describes an illustrative example of a student who was rated in the 100th percentile on his/her math score of 394 in the third grade. New York’s growth model predicted that he/she would score 399 in math for the fourth grade. In reality, he/she scored 377 which correlates to him/her answering 64 questions correctly and two questions incorrectly. As to this student, petitioner received a growth score of 27/100 based on the difference in this student’s actual test result compared to his/her predicted test result. — 895 — Petitioner’s school principal provided a reply affidavit in further support of the petition. Therein, the principal notes that petitioner’s students met or exceeded state standards at over twice the average for the past two years. The principal also discussed her personal observations of the substantial growth in achievement for all of petitioner’s students during the 2013-2014 school year. Petitioner also provided the court with a reply affidavit from professor Darling-Hammond of Stanford University. Therein, the professor reiterates the particular measurement problems for students who are either advanced in their learning or who are struggling. She again discusses the potential biases this type of model potentially inflicts upon teachers working with the lowest-performing or highest-performing classes. She also discusses how the wild swings in ratings performances for teachers across the state are emblematic of a flawed design and implementation of respondents’ growth model. Professor Amrein-Beardsley also provided the court with a reply affidavit. Therein, the professor reiterates that a teacher teaching a reasonably similar class with statistically similar absolute student performance should not experience such a notable drop as petitioner did here. She notes that respondents have not addressed the precipitous swing petitioner experienced from 14/20 to 1/20. The professor also stresses that it would have been impossible for petitioner to know what she would have needed to do to achieve any particular score in New York’s growth model. Petitioner provided the court with a reply affidavit from Dr. Burris. Therein, she points out differences in teacher ratings in several school districts from 2012-2013 to 2013-2014. More specifically, she measures changes in how many teachers were rated ineffective and how many were rated effective between the two years. Among her examples were the Scarsdale School District where 0% of teachers were ineffective in 2012-2013 as compared to 19% in 2013-2014. She also notes that Scarsdale’s effective ratings were 13% in 2012-2013 as opposed to 0% in 2013-2014. Dr. Burris also challenges the contents of Mr. Sherman’s submission. Specifically, she appears to contend that Sherman merely provided the court with graphs, charts and statistics without any meaningful data or other information to support his conclusions. Professor Pallas has also provided the court with a reply affidavit. His reply affidavit includes revised credentials, notably — 896 — that he was a member of a blue-ribbon panel convened, by New York’s Board of Regents in 2015 to discuss issues involving teacher evaluations. He contends that respondents have ignored his analysis and professional opinion that the growth model is fundamentally flawed based on its predetermined set percentage of teachers who are categorized as ineffective regardless of their actual effectiveness and their students’ achievements. He also reiterates the lack of transparency as to what a teacher needs to do to achieve any particular growth score or rating. Specifically, he again notes that at no point during the school year would a teacher know what level of student performance is associated with any particular growth score or rating. The professor contrasts this absence of information with the requirements of Education Law § 3012-c. Petitioner also provided the court with an affidavit from Dr. Sean Patrick Corcoran from New York University. Dr. Corcoran is an associate professor of educational economics. He opines that VAMs are highly variable from year to year and are sensitive to the scale of the test used to calculate them. Dr. Corc-oran also points out that VAMS are influenced by classroom, school and student factors that are beyond the teacher’s direct influence. He notes that he is unaware of any published research that claims that VAMs based on a single year of test scores are reliable. Petitioner provided the court with an affidavit from professor Jesse Rothstein of the University of California, Berkeley. Professor Rothstein was also a member of a blue-ribbon panel convened by New York’s Board of Regents to discuss issues involving teacher evaluations in 2015. He indicates that he was struck by respondents’ failure to defend the growth model’s handling of a teacher whose scores in prior years are at the top of the test score range. Professor Rothstein also stresses that effective VAM results are dependent upon averaging a teacher’s outcomes over several years so as to construct a summary measure that is less affected by determinants of student outcomes in any particular year. He also echoes Dr. Corcoran’s conclusion regarding the absence of any literature supporting the claims that VAMs based on a single year of test scores are reliable. Arbitrary and Capricious In order to find State actions arbitrary and capricious, the court must determine that respondents’ actions here were “taken without sound basis in reason or regard to the facts” — 897 — (see Matter of Peckham v Calogero, 12 NY3d 424, 431 [2009]). The burden of establishing the arbitrary and capricious standard rests with the petitioner (see Matter of Johnson Elec. Constr. Corp. v New York State Dept. of Transp., 124 AD3d 1199, 1200 [3d Dept 2015]). The court begins its analysis by acknowledging the significant budgetary concerns impacting the State’s educational system and the State’s admirable and considerable efforts to implement testing requirements in a manner that would comport with requirements for hundreds of millions of dollars of federal educational aid, while simultaneously satisfying a wide variety of competing public interests and constituents. The court also acknowledges the significant intellectual efforts and labor behind the development of the growth model system utilized by Mr. Sherman’s company and New York State. Indeed, the court fully recognizes that it does not have the educational background, resources, or time to propose a meaningful replacement for New York’s growth model system or to propose sound fixes for any arguable flaws of said system, and as such does not lightly enter into a critical analysis of this matter. Regardless, the court is constrained, on this record, to conclude that petitioner has met her high burden and established that petitioner’s growth score and rating for school year 2013-2014 are arbitrary and capricious. The court’s conclusion is founded upon: (1) the convincing and detailed evidence of VAM bias against teachers at both ends of the spectrum (e.g. those with high-performing students or those with low-performing students); (2) the disproportionate effect of petitioner’s small class size and relatively large percentage of high-performing students; (3) the functional inability of high-performing students to demonstrate growth akin to lower-performing students; (4) the wholly unexplained swing in petitioner’s growth score from 14 to 1, despite the presence of statistically similar scoring students in her respective classes; and, most tellingly, (5) the strict imposition of rating constraints in the form of a “bell curve” that places teachers in four categories via predetermined percentages regardless of whether the performance of students dramatically rose or dramatically fell from the previous year. In response to petitioner’s overwhelming submissions of evidence from learned experts in various relevant fields, respondents largely failed to adequately address the rationality or necessity of the implementation of the predetermined “bell curve.” Also unaddressed were the potential VAM biases — 898 — against teachers with high-performing students. Critically, respondents also failed to meaningfully address how petitioner’s score could have so precipitously dropped from 14 to 1 (reflecting a drop of two levels from the second-highest level of effective all the way to the lowest level of ineffective) in a single year with statistically similar scoring students. Respondents’ reliance on the “bell curve” distribution of 100 similarly situated teachers with similarly situated students lacked adequate detail for the court to enter into a reasoned analysis of same. While the court recognizes the difference in measuring student testing achievements versus student testing growth, respondents inadequately explained petitioner’s wild, statistically significant swing from 2012-2013 to 2013-2014 based on the statistics provided in this record. Finally, although respondents’ submissions attempted to explain that petitioner’s students performed: (1) less well than other similar students; and (2) failed to grow as much as predicted, they did not explain how petitioner, on this record, could have received the second lowest possible growth score and a rating that correlates to a finding of “well-below State average for similar students.” In sum, the court finds that petitioner met her burden of establishing that her growth score and rating for 2013-2014 was indisputably arbitrary and capricious. Unsealing Motion The court has not been persuaded as to the necessity of unsealing any of the confidential records at issue. Further, the court finds insufficient basis to grant leave to file an amended petition in light of: (1) petitioner’s success as to her first category of requested relief; (2) the court’s mootness finding as to the second category of requested relief; and (3) the summary nature of the proceeding. Finally, the court’s analysis of the petition did not include theories improperly raised for the first time in petitioner’s reply papers. Based upon the foregoing it is hereby ordered and adjudged that the petition is granted to the extent that petitioner’s growth score and ineffective rating for school year 2013-2014 are hereby vacated and set aside; and it is further ordered and adjudged that the petition is otherwise denied as moot and the remaining relief requested is likewise denied as moot; and it is further ordered and adjudged that based on the court’s mootness finding, the court declines to render a declaratory judgment of any kind as to petitioner’s second category of requested — 899 — relief; and it is further ordered and adjudged that petitioner’s unsealing motion is hereby denied in its entirety. Footnotes [2] Petitioner’s 2014-2015 growth score was significantly closer to 14 than 1. Additionally, her 2014-2015 growth score is back to the effective level. Thus, she dropped two full categories down from 2012-2013 to 2013-2014 and jumped two categories back up in 2014-2015. [3] More specifically, respondents advise that the scores will be used for advisory purposes only. Further, the new regulations prohibit the use of such scores for employment consequences, including teacher improvement plans, principal improvement plans and tenure determinations. Additionally, the scores will not be reflected in any teacher’s employment record. [4] The court notes that the professor also stressed the importance of having an appeal procedure in place for teachers. It should be noted that said procedure is now in place as a result of the new regulations established by the Regents.