Index No.

11381311 0

SUPREME COURT OF THE STATE OF NEW YORK COUNTY OF NEW YORK

MICHAEL MULGREW, as President of the UNITED FEDERATION OF TEACHERS. Local 2, American Federation of Teachers, AFL-CIO, on behalf of all represented employees in the City School District of the City of New York,

Petitioners,

- against-

BOARD OF EDUCA nON OF THE CITY SCHOOL DISTRICT OF THE CITY OF NEW YORK, and JOEL L KLEIN, as Chancellor of the City School District of the City of New York,

Respondents,

For a Judgment Pursuant to Article 78 of the CPLR and for Declaratory Relief Pursuant to CPLR 3001.

MEMORANDUM OF LAW IN OPPOSITION TO THE PETITION

MICHAEL A. CARlJOZO

Corporation Counsel of the City a/New York Attorney for Respondent

100 Church Street

New York, N. Y 10007

Of Counsel: JESSE I LEVINE Tel: (212) 442~3329

Matter No. 2010-004539

SUPREME COURT OF THE STATE OF NEW YORK COUNTY OF NEW YORK

~---------------------------------------------------------------------- x

MICHAEL MULGREW, as President of the UNITED FEDERATION OF TEACHERS, Local 2, American Federation of Teachers, AFL-CIO, on behalf of all represented employees in the City School District of the City of New York,

MEMORANDUM OF LAW IN OPPOSITION TO THE PETITION

Petitioners,

Index No. 113813/10 (Justice Kern)

-against-

BOARD OF EDveA nON OF THE CITY SCHOOL DISTlUCT OF THE CITY OF NEW YORK, and JOEL I. KLEIN, as Chancellor of the City School District of the City of New York,

Respondents,

For a Judgment Pursuant to Article 78 of the CPLR and for Declaratory Relief Pursuant to CPLR 3001.

------..,.----------- .... ----------------- ... x

ST ATEMENT OF THE CASE

Respondents Board of Education of the City School District of the City of New York

(referred to herein as "DOE") and Joel L. Klein, (referred to herein as the "Chancellor") as

Chancellor of the City School District of the City of New York (the "District"), submit this

Memorandum of Law in opposition to the Petition herein and petitioner'S motion for a

preliminary injunction,

Petitioner in this Article 78 proceeding seeks to enjoin respondents from complying with

requests for documents made by several news organizations pursuant to the New York Sate

Freedom of Information Law (Public Officers Law §84 et seq., "FOIL"). Those documents

constitute in their totality a statistical tabulation of data regarding teacher and student

performance in grades four and eight in New York City's Public Schools. They are in the form of

Teacher Data Reports (''TDR''s). They are objective data of individual teacher performance, and

they rank teachers on a curve in comparison with their peers. The critical issue is whether the

TORs must be disclosed under FOIL with the names of individual teachers included.

This proceeding comes before the Court in an unusual procedural posture. As stated in

Capital Ncwspa~rs Div. of Hearst Corp. v. Bums, 67 N.Y. 2d 562, (1986), there is no statutory

authority for a "reverse FOIL" proceeding by a party seeking to prevent disclosure of records

under FOIL, and FOIL is permissive in nature so that the agency involved could disclose the

records without a FOIL request.'

However, as the records were intended for internal use, and a representative of DOE had

agreed with Petitioner that DOE would raise appropriate objections to any FOIL request for the

TOR's, 2 Respondents will not, in this proceeding only, raise any defense to the petition based on

standing. We believe that the case law interpreting FOIL requires DOE to disclose the

documents requested. It is anticipated that the Press Intervenors will present arguments on these

issues to the Court for consideration.

I The issue of intervenor's standing to prosecute this appeal is not before us, however, and the parties below consented to Tuffey's intervention to protect his interests in nondisclosure. Thus, on the record before us we do not reach the issue whether a public employee has a cause of action under FOIL to prevent disclosure (compare, Carpenter v. City of Plattsburgh, 105 A.D.2d 295 [Main, J.], affd for reasons stated below 66 NY2d 791.).

2 "Respondent's long-standing promise of confidentiality to the intervenors is irrelevant to whether the requested documents fit within the Legislature's definition of "records" under FOIL.". Washington Post Co. v. New York State Ins. Dep't. 61 N.Y.2d 557, 565 (N.Y. 1984), See also Anonymous v, Board of Educ., 162 Misc. 2d 300, 303 (N.Y. Sup. Ct. 1994) ("[i)n any event, an agreement to keep secret that to which the public has a right of access under article 6 of the Public Officers Law would be unenforceable as against public policy").

2

FACTUAL BACKGROUND

DOE's Teacber Data Initiative

As set forth in the Affidavit of Joanne Cannon, Deputy Executive Director of DOE's

Research and Policy Division, submitted herewith, the documents at issue were created pursuant to a special initiative implemented by respondents to measure and improve teacher performance in the classroom through value- added modeling ..

During the 2006-07 school year, the DOE began work to provide data to schools about their teachers using value-added statistical calculations (subsequently named the Teacher Data Initiative, TDI). Value-added data were first shared with approximately 100 schools in a pilot program during the 2007-08 school year. During the 2008-09 and 2009-10 school years, the DOE provided value-added information to all schools with available data. The data that are the subject of the UFT's petition are those shared during the 2009-10 school year ..

The pilot program data and the data released during the 2008-09 school year were calculated by an external vendor, the Battelle Institute. The value-added calculations released during the 2009-10 school year were conducted by the Value-Added Research Center (V ARC). In 2009 the DOE sought vendors through a competitive bidding process. The DOE switched vendors because the term of the contract with Battelle had ended and because of the requirements of the competitive bidding process, not because of dissatisfaction with the Battelle Institute (who did not bid as part of the competitive bidding process). Ultimately; the DOE awarded the project to V ARC, which is located within the School of Education at the University of Wisconsin at Madison.

V ARC is a leader in the field of educational research and statistical modeling. It receives funding from the United States Department of Education, the National Science Foundation, and the National Center for Education Statistics, and it partners with school districts around the

3

country to develop statistical models for measuring the performance of students, teachers, schools, and school districts.

V ARC has developed a "value-added" statistical model for New York City public school math and English Language Arts (ELA) teachers in grades 4-8. The V ARC model ("the model") was developed in coordination with DOE, with input from teachers, principals, administrators, researchers, and from representatives of the United Federation of Teachers.

"Value-added" is a statistical term used to refer to a family of statistical models that calculate the effects of individual teachers or schools on student test performance. The VARC value-added model developed for New York City was based on standard norms and the most current research in the fields of mathematics and statistics concerning how to calculate valueadded data. Value added modeling is utilized by school districts and in states across the country to measure teacher and school performance, including in Chicago, Washington D.C., Charlotte, Tennessee, Ohio, Texas, and Pennsylvania. New York State recently passed a law requiring the use of value-added in teacher evaluations.

VARC's value-added model is calculated using data on all students tested on the New York State math and ELA assessments in grades 4-8 over multiple years. Included in the model is demographic and performance information (also known as predictor variables) on every student tested. Using this information accounting for how students with different background characteristics performed on that year's New York State math and ELA tests, the model generates the predicted performance on these tests for students in each teacher's class Conceptually, value-added is the difference between the average predicted score of a teacher's students and the average actual score of these students. The value-added score represents the teacher effect - the average student performance on the tests subtracting the impact of factors

4

outside of a teacher's control. Teachers whose students out-perform predictions have positive value-added scores; teachers whose students under-perform predictions have negative valueadded scores.

The students who are included in a given teacher's value-added calculations represent those students the teacher was responsible for teaching in math or ELA. If two teachers shared responsibility for teaching math or ELA to a group of students, the pair of teachers (also known as co-teachers) would receive one value-added score. Only students who were in a teacher's classroom for the full school year would be included in that teacher's value-added calculations. Similarly, value-added scores would only be calculated for teachers who taught for the full school year.

The model incorporates the following individual student characteristics as factors in

calculating a student's predicted score on the tests:

• The student's prior year ELA and math score;

• Whether the student receives free or reduced price lunch;

• Whether the student has disabilities, and if so, what kind of special education services the student receives;

• Whether the student is a current or former English Language Learner

• The number of suspensions a student has received, if any, for the prior

year;

• The number of student absences for the prior year;

• Whether the student was retained in his/her grade for the prior year;

• Whether the student attended summer school during the prior summer;

• Whether the student is new to the current school;

5

• The student's ethnicity;

• The student's gender.

In addition to these individual student characteristics, the characteristics of a particular classroom in which a student is situated are also incorporated into the model to generate a student's predicted score on the tests. The following classroom characteristics are incorporated into the model:

• A classroom's average prior-year score on the math and English language assessment tests

• The percent of students in the classroom receiving free vs. reducedprice lunch;

• The percentage of students in the classroom who receive different types of special education services;

• The percentage of students in the classroom who are current and former English Language Learners;

• The average number of suspensions in the prior year for students in the classroom;

• The average number of absences in the prior year for students in the classroom;

• The percentage of students in the classroom who were retained in the grade for the year prior;

• The percentage of students in the classroom who attended summer school during the prior summer;

• The percentage of students who are new to the school;

6

• The ethnic make-up of the classroom;

• The gender make-up of the classroom;

• The class size.

The model generates an overall value added score for the teacher for each subject and grade. Value-added scores are calculated for the most recent school year of information, as well as a cumulative value-added score using data for the last four school years where available. Value-added score are also calculated measuring a teacher's effectiveness for various subgroups of students.

For the value-added data shared with schools during the 2009-10 school year, the valueadded scores, in addition to the predicted and actual scores, were reported in" proficiency ratings," This is the same scale upon which student test scores are reported in NYC.

V alue-added scores are also presented as percentile rankings. The percentile ranking simply ranks all the value-added scores for teachers in a particular grade and subject with similar years of experience from highest to lowest - the teacher with the highest value-added score is the 99th percentile, the teacher with the lowest value-added scores is in the 0 percentile. DOE uses four levels of experience when making this comparison: first-year teachers, second-year teachers, third-year teachers, and teachers with four or more years of experience. The purpose of the percentile score calculations is to enable one to compare a teacher's value-added score to other teachers with similar experience levels.

The value-added model also computes "confidence intervals" (also known as ranges) around the percentile score. A confidence interval is a statistical term and calculation. Confidence intervals can be calculated from all inferential statistics and it is common practice to report them with the results of statistical calculations, The DOE chose to report confidence

7

intervals in the Teacher Data Initiative to be transparent about their calculations and the inferences that can be drawn from them.

A confidence interval can be calculated with different confidence levels, the most common of which is a 95% confidence level. A 95% confidence level is used in the NYC valueadded model, The 95% confidence level can be interpreted to mean that one can be 95% confident that the teacher's value-added score is within that range, most likely at or near the value reported. Confidence intervals reflect the fact that all statistical calculations are based on a sample of data. For example, one would expect that one would have more confidence in a calculation based upon 200 students than one based upon 20 students. Indeed, confidence intervals are generally larger-that is, indicative of greater uncertainty-when a teacher has taught for fewer years/ has fewer students or when the value-added score is in the middle of the distribution of scores. Confidence intervals are also greater the weaker the relationship is between the predictor variables and the outcome (in the current model, student performance). It is in part for this reason that we chose to include a number of predictor variables in our model.

All of the above calculations and data are stored in a computer database housed at V ARC. Depending on the desired output, the data can be collated and displayed in various formats. One such format is known as a Teacher Data Report ("TDR"). A teacher can access hislher TOR via internet once the IDR has been released for a given school year. Similarly, a principal can access the TORs for hislher teachers as well as a school summary report via internet once the TDRs have been released for a given school year.

The model itself does not contain any subjective weighting or relational elements as petitioner asserts. To the contrary, the model's calculations are purely objective. No subjective decisions are made by any person at DOE with respect to the weight that each variable is

8

accorded when determining a student's predicted score on a test. Rather, the formula itself mathematically determines the degree to which various student characteristics actually do impact student performance, while holding all other variables constant.

Each year reports are released, the DOE has worked to continually improve the model and the reports, respond to feedback (from researchers, principals, teachers, administrators, and the UFT), and make the data more useful to New York City principals and teachers. DOE made changes to the reports released in 09-10 from the prior year to make them easier to read and interpret. For example, it eliminated citywide percentile scores and tables summarizing other sections of the report since they provided little new information, making it easier to focus on critical results. In addition, when DOE changed vendors, it also made improvements to the value-added model. For example, DOE had previously only differentiated between students who did and did not have disabilities. DOE changed this for the reports released in 09-10 to differentiate between the special education services students with disabilities received. DOE also increased the number of students a teacher must have to generate a value-added report. Previously the minimum number of students was 6. DOE changed it to 10 students for all grades in math and elementary ELA, and 20 students for ELA middle school grades.

Moreover, unlike subjective assessments by peers or supervisors, the model provides objectively verifiable results. Every year, the DOE engages in a verification process with schools where they can make changes or corrections to the data linking teachers to students (i.e., teacher's classroom rosters) that is used to generate these reports. The verification process for the value-added data released during the 2009-10 school year took place in the spring of 2009. Principals received weekly reminders via the Principals Weekly email bulletin about this verification process. The verification process was described on the Principal Portal webpage,

9

was announced in the school data specialist bulletin and at data specialist training sessions, and was also announced on the Teacher Page, a website for teachers. In addition, the DOE informed the UFT about the verification process.

While errors in the TORs can arise in instances where schools omit information or fail to verify student-teacher assignment data, DOE provides support for principals when a possible error is brought to his/her attention and allows for data to be corrected in the following year's report. To that end, the DOE has a helpdesk for principals and school staff engaged in the verification process (the Accountability Helpdesk), which provides schools with guidance about how to verify data, interpret the reports and about what to do in the event they believe an error exists in the student-teacher assignment data on a TOR. In the event an error is identified, principals can correct the information, and DOE will update the information so that it is reflected on the following year's report. Any reports disclosed pursuant to this proceeding will reflect the most recent verification process.

FOIL Requests for Teacher Data and DOE's Responses

As set forth in the accompanying affirmation of Joseph Baranello, Central Records

Access Officer of DOE, DOE has previously released much of the data collected as a part of the TO! in response to FOIL requests received by DOE beginning in early 2009. DOE determined that release of the data was proper, pursuant to New York State Freedom of Information Law, subject to certain redactions, as explained more fully below. DOE did not receive any FOIL requests for the release of any teacher data containing individual teacher names until August 16, 2010. Prior to that time, all FOIL requests for teacher data collected as a part of DOE's Teacher Data Initiative either explicitly or implicitly requested teacher data without individual teacher

names.

FOIL Request #F6322

10

By email dated July 2, 2009, DOE received a request for teacher data from Ms. Jennifer Medina, education reporter at the New York Times. Specifically, Ms. Medina requested "[a]l! teacher data reports for the 2007-2008 and 2008-2009 school year, with corresponding information about teacher performance, grade level, school, and geographic district." Ms. Medina's request (assigned an internal DOE reference number of F6322) did not include a request for individual teacher names.

By letters dated August 6, 2009 and October 2, 2009, Mr. Baranello informed Ms.

Medina that due to the volume and complexity of the requested records and in light of the need to determine which of the records, if any, were subject to redaction pursuant to federal statute, DOE required additional time in order to properly respond to FOIL request F6322.

After an administrative appeal DOE responded to FOIL request F6322 by letter December 8, 2009, by providing the requested data for the 2007-2008 school year, with the following redactions:

• school name and school number;

• totals of numbers of students;

• student prior proficiency ratings; and

• all student sub-group information.

In addition, teacher file numbers were encrypted, in order to prevent unlawful access to a teacher's confidential information contained in DOE's personnel databases. As FOIL Request F6322 did not contain any request that a release of records include individual teacher names, no names were provided.

FOIL Request #F6369

11

DOE subsequently received a request for teacher data from Meredith Kolodner at the New York Daily News by letter dated July 27, 2009 (F6369). Specifically, Ms. Kolodner requested "access to and if necessary copies of teacher data reports produced between 911108 and 8/1/09."

By letter dated July 27, 2009, DOE responded to Ms. Kolodner's request by providing

the requested data contained within the TDRs, with the following redactions:

• school name and school number;

• totals of numbers of students;

• student prior proficiency ratings; and

• all student sub-group information.

Teacher file numbers were again encrypted, in order to prevent unlawful access to a teacher's confidential information. As FOIL Request F6369 did not contain any request that a release of records include individual teacher names, no names were provided.

FOIL Request #F6544

By letter dated November 4, 2009, DOE received a request for teacher data from Kimberly Liu at the NAACP Legal Defense and Educational Fund, Inc. ("LDF"). Ms. Liu enumerated twenty-two (22) separate requests for, inter alia, information relating to implementation of the Teacher Data Initiative, methodology and procedures of data collection, and data results sorted according to various student and teacher categories.

By letter dated April 12,2010, Baranello provided a partial response to Ms. Liu's request by providing teacher data for the 2007-2008 school year the with the following redactions:

• school name and school number;

• totals of numbers of students;

12

• student priority proficiency ratings [see above]; and

• all student sub-group information.

Teacher file numbers were again encrypted, in order to prevent unlawful access to a teacher's confidential information. As FOIL Request F6544 did not contain any request that a release of records include individual teacher names, no names were provided.

After an administrative appeal, DOE provided LDF with teacher data collected during the 2007-2008 school year, redacted in accordance with the decision on appeal.

FOIL Request #F6850

By letter attached to an email dated April 7,2010, Ms. Medina of the New York Times again requested TDR data pursuant to FOIL (F6850). Specifically, Ms. Medina requested "[a]ll teacher data reports for the 2008-09 school year at the school and district level, with all relevant underlying information." FOIL Request F6850 did not contain a request for individual teacher

names.

As Ms. Medina's second request was received prior to the above appeal by LDF and subsequent decision by Mr. Best, DOE responded to Ms. Medina's request by providing teacher data for the 2007-2008 school year with redaction of information consistent with the redactions made in FOIL responses F6322, F6369, and F6544.

None of the above FOIL requests (F6322, F6369, F6544, F6850) contained a request for teacher names. The New York Times explicitly stated that teacher names were not being requested; the two remaining requesters implicitly indicated as much by omitting any mention of teacher names in F6369 and F6544. Moreover, in a recent FOIL request for later years of teacher data (F6319), LDF stated that they are not seeking student names. Accordingly, names of individual teachers were not provided with TDI data because they would have constituted

13

non-responsive material.

Moreover, because the above requests did not seek disclosure of

individual teacher names, DOE made no determination in connection with those requests

regarding whether disclosure of teacher names was required pursuant to New York State's

Freedom of Information Law.

Requests that Give Rise to the Instant Petition

Beginning on August 16,2010 and continuing through October 27,2010, DOE received

nine (9) FOIL requests which counsel informs me give rise to the instant Petition. All of the

above FOIL requests (F7098, F7105, F7227, F7252, F7272, F7273, F7274, F7275, and F7304)

explicitly request that a release of records include individual teacher names. All of the above

FOIL requests have been acknowledged as received by DOE, but no records have been provided

in response, as a consequence of the pending litigation.

THE STATUTORY FRAMEWORK

New York's FOIL is designed to provide transparency in government

operations for the benefit of the public, Where appropriate, government agencies may withhold

information based on statutory exemptions from disclosure.

The legislative intent is clear. Public Officers Law provides:

§84. Legislative declaration. The legislature hereby finds that a free society is maintained when government is responsive and responsible to the public, and when the public is aware of governmental actions. The more open a government is with its citizenry, the greater the understanding and participation of the public in government.

As state and local government services increase and public problems become more sophisticated and complex and therefore harder to solve, and with the resultant increase in revenues and expenditures, it is incumbent upon the state and its localities to extend public accountability wherever and whenever feasible.

The people's right to know the process of governmental decision-maktng and to review the documents and statistics leading to determinations is basic to our society. Access to such information should not be thwarted by shrouding it with the cloak of secrecy or confidentiality. The legislature therefore declares that government is the public's business and that the public, individually and

14

collectively and represented by a free press, should have access to the records of government in accordance with the provisions of this article.

Basic Provisions of FOIL

All agency records must be disclosed on request to any member of the public, except records which fall into one or more of ten specified exemptions, set forth in Public Officers Law ("POL") §87(2).

There is a presumption in favor of disclosure, which must be overcome by an agency if it wants to withhold records.

FOIL does not require that an agency withhold any records, even if they are covered by one or more exemptions. An agency can always choose to disclose records under FOIL.

POL Section 87 subd. 2 contains the provisions that exempt certain material from disclosure. Those provisions relevant here are as follows:

"Each agency shall, in accordance with its published rules, make available for public inspection and copying all records, except that such agency may deny access to records or portions thereof that:

... (b) if disclosed would constitute an unwarranted invasion of personal privacy under the provisions of subdivision two of section eighty-nine of this article; [or]

... (g) are inter-agency or intra-agency materials which are not: i. statistical or factual tabulations or data:

ii. instructions to staff that affect the public; iii. final agency policy or determinations;

iv. external audits, including but not limited to audits performed by the comptroller and the federal government(.]"

POL section 89 (2)(b), referenced in section 87(2)(b), provides in relevant part as follows:

"An unwarranted invasion of personal privacy includes, but shall not be limited to:

i. disclosure of employment, medical or credit histories or personal references of applicants for employment[.}"

15

ARGUMENT

Petitioner challenges DOE's determination that the TORs are statistical tabulations' that

must be disclosed and that the personal privacy exemption is inapplicable here. The challenge

should be denied. DOE correctly determined that the TDRs are statistical tabulations that are not

exempt from disclosure; moreover, the personal privacy exemption does not apply.

In Capital Newspapers Div. of Hearst Corp. v. Burns, 67 N.¥. 2d 562 ,566-567

( 1986) where the intervenor sought to prevent the disclosure of [certain] records, the Court

stated:

Exemptions are to be narrowly construed to provide maximum access, and the agency seeking to prevent disclosure carries the burden of demonstrating that the requested material falls squarely within a FOIL exemption by articulating a particularized and specific justification for denying access (see, Matter of Farbman & Sons v New York City Health & Hosps. Corp., 62 NY2d 75, 80, supra; Matter of Fink v Lefkowitz, 47 NY2d 567,571, supra).

Petitioners cannot satisfy the heavy burden they bear to demonstrate that DOE's determination

was in error.

THE TDRS TABULATIONS DISCLOSED IN REQUEST

POINT I

ARE STATISTICAL

WHICH MUST BE

RESPONSE TO A FOIL

As evidenced by its legislative declaration, New York's Freedom ofInfonnation Law

was enacted in order to assist the public in understanding the operations of its government and

3

For purpose of this analysis, DOE assumed that the materials were inter- or intra-agency materials. Such documents would be exempt from disclosure unless they constitute, among other things "statistical or factual tabulations or data." See POL § 87(2)(g)(i).

16

the work of public employees. Public Officers Law § 84 provides that "[tjhe people's right to know the process of governmental decision-making and to review the documents and statistics leading to determinations is basic to our society." FOIL does not distinguish between types of statistical methodology that mayor may not be exempt from release, nor does it exclude from release entire categories of statistics of the sort that the UFT attempts to argue (in the face of evidence to the contrary) are "subjective." Rather, FOIL's is intended to advance the public's interest in accessing data, and in particular the type of data found in teacher data reports, which measures the effectiveness of a group of public employees and which have been calculated using accepted statistical methodologies.

Public Officers Law § 87 (2) (g), even when it applies, "permits access to records or portions thereof which contain any statistical or factual information, policy or determinations upon which the agency relies" (citation omitted) .... Factual observations are not exempt from disclosure, even in documents issued before final decision (citations omitted) New York 1 News v. Office of the President, 231 A.D. 2d 524 (2d Dept. 1996)

Consistent with the requirements of FOIL, DOE determined that the docwnents, with the teacher names included, were not exempt from disclosure. As fully detailed in the Cannon affidavit, the TDRs consist of statistical tabulations. DOE, working with an outside vendor, identified a set of variables applicable to determining the effect of a teacher on the academic improvement of a child. Typical with statistical analysis, the formula is designed to account for factors for which the teacher cannot control, but which have been determined to impact on performance, such as poverty. disabilities, and disciplinary history. The data about particular students is then entered into the formula to reach a conclusion. This is the essence of a statistical analysis.

17

Petitioner asserts that only "objective information" reflecting "objective reality" comes

within the definition of statistical or factual tabulations or data. In doing so, it erroneously relies

on what it asserts is a definition of the term "statistical or factual tabulations or data" set forth in

9 New York Code of Rules and Regulations ("NYCRR") §14.l and on dictum of the Court of

Appeals in Gould v. New York City Police Department, 89 N.Y.2d 267 (1996).

9 NYCRR §14.1 defines terms used in Part 14 of9 NYCRR, which applies to records of

the Governor's Executive Chamber.4 What Petitioner presents as the definition of a single term,

"statistical or factual tabulation," is actually two definitions of separate terms: "statistical

tabulation" ("a collection or orderly presentation of numerical data logically arranged in columns

and rows or graphically") and "factual tabulation" ("a collection of statements of objective

information logically arranged and reflecting objective reality, actual existence or an actual

occurrence"). The definition of "statistical tabulation" makes no reference to the objectivity or

provenance of the numerical data presented; the definition of "factual tabulation" does not

indicate that the statements collected are primarily (or at all) numerical or statistical?

In Matter of Dunlea v. Goldmark, 54 A.D.2d 446 (3d Dept. 1976), affd on opinion

below, 43 N. Y.2d 754 (1977), one of the seminal cases in determining the treatment of statistical

and factual material under FOIL, the Court recognized the distinction between statistical

tabulations and factual tabulations, There, the Court held that certain budget examiners' files and

worksheets, consisting primarily of numerical notations and tabulations, notwithstanding that

4 The rules of the Committee on Open Government, set forth in 21 NYCRR Part 1401, are generally applicable to agencies subject to FOIL, but do not defme "statistical or factual tabulation" or any similar term. Moreover, whatever persuasive value petitioner may claim for section 14.1 disappears upon closer scrutiny.

5 See also, 9 NYCRR 6150.2, defining for the Division of Criminal Justice Services statistical and factual tabulation in similar fashion.

18

such materials might contain approximations, estimates or hypotheses, were disclosable under

FOIL.

Discussing the meaning of the statutory term "statistical or factual tabulations", the Court referred to the above-quoted definitions and noted that "[t]his definition [of "factual tabulation"] attempts to restrict [the term] to 'objective information' or 'objective reality'." The Court stated, however, that "there is no statutory requirement that such data be limited to 'objective' information and there is no apparent necessity for such a limitation", and concluded that "[t]he appellants have not demonstrated that the work sheet is not to be public information." Id. at 448- 449.

Similarly, the Appellate Division, Third Department required the disclosure of field auditors' work papers created in an audit; in Matter of Polansky v. Regan, 81 AD.2d 102 (3d Dept. 1981). The Court, noting its ruling in Matter of Dunlea v. Goldmark, held that "[t]he mere fact that some of the data might be an estimate or a recommendation does not convert it into an expression of opinion or naked argument for or against a certain position and such tabular data is not excluded from disclosure." 81 AD.2d at 104. The Court further held that certain "explanations" or "remarks" contained in the field auditors' work papers "could be in the nature of opinion or argument for or against certain matters being considered in the audit process" and could therefore be redacted. Id.

As Petitioner notes at page 13 of its Memorandum, the Court of Appeals, in Gould v.

New York City Police Department, states that "[ f]actual data ... simply means objective information, in contrast to opinions, ideas, or advice exchanged as part of the consultative or deliberative process of government decision making." 89 N. Y.2d at 277 (citations omitted). However, it is clear that the Court's statement in dictum was not intended to draw a bright line

19

around "objective information" as the only material disclosable as "statistical or factual

tabulations or data". Rather, it is an attempt to delineate the real focus of the Court's interest,

"opinions, ideas, or advice exchanged as part of the consultative or deliberative process of

government decision making", by indicating what stands in sharpest contrast. That the Court did

not intend its statement to be taken literally is clear from its holding, in the same case, that

witness statements collected by the Police Department were disclosable as "a factual account of

the witness's observations", even though, as the Police Department pointed out, "there is no

assurance of the statements' accuracy and reliability." Matter of Gould v. New York City Police

Department, 89N,Y.2d 267, 277.

It is not necessary for the Court to determine whether TDRs are subjective or objective,

as the case law makes it clear that that is not the determinative factor, The TDRs fit within the

analysis of statistical tabulations undertaken by the courts. To the extent that decisions are made

as to what are the appropriate variables, there is of course human decision making involved in

the determination of the variables to be included. As the list of variables evidences, however,

DOE has selected all possible variables that can be consistently accounted for." Once the

variables are selected, statistical analysis applies the variables against student test scores in order

to arrive at a determination of teacher impact on test scores. The variables, rather than being

subjective, support the use of the TDRs by increasing the reliability of the conclusions reached.

DOE is not simply determining teacher effectiveness by the absolute scores attained by students

6 The formula cannot account for random events such as noise outside a classroom window or the fact that a student may not feel well on the day of the test. These are truly subjective factors that are not susceptible to data analysis.

20

- such a determination would be unreliable due to a failure to account for the other factors that impact performance.

DOE's selection of and refining of the variables to which the available data is applied is a strength not a weakness, as asserted by petitioner, in the process, It does not change the essential characteristics of TDRs as statistical compilations,

The two cases primarily relied on by Petitioner to argue that TDRs are not disclosable "statistical or factual tabulations or data," Matter of David v, Lewisohn; 142 A.D.2d 305 (3d Dept. 1988), lv, to appeal denied, 74 N,Y,2d 610 (1989) and Matter of General Motors Corp. v. Town of MassenSh 180 Misc.2d 682 (Sup. Ct., Lawrence Co. 1999), are inapplicable here, Both involved determinations by tax assessors regarding whether certain real property sales were comparable, for assessment purposes, to the property transfer being taxed. In Lewisohn, the underlying sales data was ordered disclosed, but the assessor's comparability determination, indicated by a letter notation, was held exempt from disclosure as deliberative materiaL In Massena, the underlying sales data was presented as part of the comparability determination and was therefore held exempt from disclosure. In both cases, the determination was based entirely on the non-quantifiable training and judgment of a single person, and was therefore clearly subjective and deliberative in nature. In like manner, the subjective written observations of a teacher by a single supervisor are exempt from disclosure. Such observations are purely SUbjective assessments.

The results of the TDRs, by contrast, are derived from the unvarying application of numerical data according to a specified formula that reflects the combined input of many

21

individuals. Unlike the facts of Lewisohn and Massena, no single subjective determination can be identified in the TDRs. Again in contrast to Lewisohn and Massena, the TDRs contain no judgment or opinion regarding a particular data subject (in this case, teachers). 'The TDRs aggregate data against a defined set of variables to produce a result. Thus, they are equally as disclosable as the work papers ordered disclosed in Goldmark and Regan. See also,Matter of Professional Standards Review Council v. New York State Dept. of Health, 193 A.D.2d (3d Dept 1993) (numerical ratings of proposals recorded on procurement evaluation sheets held dis closable as "statistical tabulations").

In its discussion of Lewisohn, Petitioner's Memorandum attempts, at page 16, to analogize the compilation of real estate sales data in that case to the compilation of teacher and student performance data in TDRs ("data from New York City real estate transactions was collected and entered into the City's computers, much as the DOE aggregated student and other data in creating the TDRs"). It is clear, however, that computers in Lewisohn were only a medium for the storage of data considered by tax assessors, see 142 A.D.2d at 306, not for the transformation of data into a collection or orderly presentation of numerical data logically arranged in columns and rows or graphically, as in the TDRs. The TDRs thus constitute a statistical tabulation, not merely an assembled mass of data.

Similarly, in its discussion of Massena, Petitioner's Memorandum, at page 17, quotes a portion of the Court's opinion indicating that all factual information which serves as the basis for a subjective determination is exempt from disclosure. To the extent it appears to have reached this conclusion, the lower court erred, and its opinion is in this respect contradicted by other case law, including Lewisohn, See 142 A.D.2d at 307.

22

Other arguments invoked by Petitioner to support the non-disclosure ofTDRs, the

alleged unreliability of some underlying data and its alleged capacity to mislead the public, are

not independent factors in the analysis of the disclosability of "statistical or factual tabulations or

data". No case law supports the withholding of statistical data solely because it contains errors,

and both Goldmark and Regan point to the irrelevance of that factor. In any event, the TOR

process included a recertification process and individual claims of error are addressed by DOE.

POINT II

THE RELEASE OF THE TDRS WILL NOT CONSTITUTE AN UNW ARRANTED INV ASION OF TEACHERS' PRlV ACY

Disclosure of the TDRs will not constitute an unwarranted invasion of teachers' privacy

pursuant to POL §§87(2)(b) and 89(2)(b) and is, accordingly, required by FOIL. Petitioner

contradicts FOIL and case law in construing otherwise. Petitioner's Memorandum opens by

citing, at page I, a 2008 letter from a Deputy Chancellor of DOE to the UFT agreeing to keep the

teacher data information confidential in the event it became the object of FOIL requests.

However, an agency's agreement to maintain confidentiality is of no consequence if it is

obligated under FOIL to disclose requested information.

Public employees generally lack an expectation of privacy in information concerning

their performance of public functions. The New York State Committee on Open Government

("COG"), whose advisory opinions are accorded persuasive effect by the courts, has repeatedly

issued advice supporting the view that employees' right to privacy, when it exists at all, is far

23

more limited than that of the general public, and has advised that names of employees must be released with respect to records relevant to their job duties. See, e.g .. COG FOIL Advisory Opinion No. 16407 (January 17, 2007) (advising that FOIL required a board of education to disclose the names of employees issued computers). COG has also specifically advised DOE that in the context of yearly ratings that principals assign to teachers, FOIL requires the DOE to disclose any "statistical or factual information" contained in records of those ratings. See COG FOIL Advisory Opinion No. 14287 (October 17,2003).

Applying the principle that employees enjoy a lesser right of privacy than do members of the public, New York courts have specifically held that records relevant to the job performance of public employees (i.e., as the teacher data reports are) must be disclosed under FOIL even when they identify particular employees. See e.g .. Capital Newspapers v. Burns, 67 N.Y.2d 562 (1986) (record of sick time taken by particular officer not exempt from FOIL disclosure); Gannett Co. v. County of Monroe, 59 A.D.2d 309 (4th Dep't 1977) (approving release of "names, job titles, and salary levels" of terminated employees), affd 45 N.Y.2d 954 (1978); Farrell v. Village Board of Trustees, 83 Misc. 2d 125 (Sup. Ct. Broome Co. 1975) (authorizing disclosure of written reprimands, including the names of police officers reprimanded); Faulkner v. Del Giacco, 139 Misc. 2d 790 (Sup. Ct. Albany Co. 1988) (finding "no basis to support the claim that releasing the names of [prison] guards accused of inappropriate behavior is an unwarranted invasion of their personal privacy"); Powhida v. City of Albany, 147 A.D.2d 236 (3d Oep't 1989) (finding records related to investigation of complaint against police officer subject to disclosure under FOIL).

Petitioner's case citations in support of the assertion that TDRs (or at least the names of teachers who are the subjects of the reports) must be withheld from disclosure under the FOIL

24

exemption for unwarranted invasion of personal privacy are inapplicable. One, Matter of Bahlman v. Brier, 119 Misc.2d 110 (Rensselaer Co. 1983), holds that records of sick leave time taken by public employees are exempt from disclosure on privacy grounds. That case has not been good law since the Court of Appeals held, in Matter of Capital Newspapers v. Burns, supra, that disclosure of the sick time records of police officers did not constitute an unwarranted invasion of personal privacy. See also Matter of Scaccia v, Division of State Police, 138 A.D.2d 50 (3d Dept. 1988).

Another case, Harris v. City University of New York, Baruch College, 114 A.D. 2d 805. (I st Dept. 1985) cited by petitioners, is distinguishable from the present case. There, the petitioner, who had been denied tenure, sought the curriculum vitae of other employees who had been granted tenure. The court held that the information could be disclosed, with the names redacted, because the petitioner could conduct the desired comparison without reference to the names of his colleagues. In Harris, completely unlike this case, the information subject to disclosure pertained to pre-public employment.

Lastly, in Matter ofObiajulu v. City of Rochester, 213 A.D.2d 1055 (4th Dept. 1995), the court stated that performance evaluations were part of an employee's "employment history" and thus came within FOIL §89(2)(b )(i), which includes within the scope of the personal privacy exemption "employment, medical or credit histories or personal references of applicants for employrnentj.]" TORs are not employment histories. Further, there are several factors that distinguish them from the performance evaluations at issue in Obiajulu in ways that mandate their disclosability. Most importantly, they do not, as noted above, reflect the personal opinion of a single person regarding the subject employee, but instead represent the impersonal, objective application of a wide range of data to provide information about an employee's performance.

25

Moreover, they provide a measure of effectiveness for which significant public interest exists, and they are designed 10 be as objective and informative as possible. Indeed the TDRs contain a detailed amount of information that measures a teacher's effectiveness with respect to his or her students as a whole, as well as to sub-groups of students, subdivided by categories such as gender, ELL status, and student performance level.

As noted above, notwithstanding the holding in Obiajulu, New York courts have also specifically held that records relevant to the job performance of public employees must be disclosed under FOIL even when they identify particular employees. See e.g., Capital Newspapers v. Burns, supra. In Kwasnik v. City of New York, 262 A.D.2d 171 (l st Dept. 1999), the Court held that all prior public employment, as reflected in the resume or job application of a public employee, as well as other experience required to qualify for the employee's current position, was subject to disc1osure. The Court specifically rejected the argument of respondent City University of New York that the requested material was exempt from disclosure under FOIL §89(2)(b)(i). In Faulkner v. Del Giacco, 139 Misc. 2d 790 (Sup. Ct. Albany Co. 1988), the Court found "no basis to support the claim that releasing the names of [prison] guards accused of inappropriate behavior is an unwarranted invasion of their personal privacy". See, also, Powhida v. City of Albany, 147 A.D.2d 236 (3d Dep't 1989) (finding records related to investigation of complaint against police officer subject to disclosure under FOIL).

Similarly, COG has stated on numerous occasions that all information in a public employee's resume or job application relevant to the employee's qualification for his or her position or to the performance of his or her official duties is discJosable under FOIL. See, ~, COG FOIL Advisory Opinion No. 15116, dated January 14,2005; 10955, dated July 21, 1998.

26

As noted above, COG has advised that the DOE must disclose not only ratings that principals

assign to teachers, but also the statistical information underlying them. See COG FOIL Advisory

Opinion No. 142871 supra. Thus, the holding in Obiajulu would seem to be the exception rather

than the rule concerning materials relevant to the work of public employees.

To the extent that any invasion of privacy would occur by release of the TDRs, DOE

correctly determined that this is outweighed by the public interest in disclosure. In determining

whether the invasion is warranted, courts balance the public interest in disclosure of the

requested information against the privacy interest of the individual subject or subjects. Thus, in

Empire Realty Corn:. v. New York State Div. of the Lottery, 230 A.D.2d 270, 273 (3d Oep't

1997). the Court said, "[w]hat constitutes an unwarranted invasion of personal privacy is

measured by what would be offensive and objectionable to a reasonable [person] of ordinary

sensibilities .... This determination requires balancing the competing interests of public access

and individual privacy" ( Matter of Dobranski v Houper, 154 AD2d 736, 737 [citation omitted];

see, Matter of Ruberti, Girvin & Ferlazzo v New York State Div. of State Police, 218 AD2d

494). "

The balancing test required in the application of the personal privacy exemption has been

articulated by courts in a number of cases, See also N.Y, Times Co. v. City of New York Fire

Dep't, 4 N.¥. 3d 477, 485 (N.Y. 2005):

The recognition that surviving relatives have a legally protected privacy interest, however, is only the beginning of the inquiry. We must decide whether disclosure of the tapes and transcripts of the 911 calls would injure that interest, or the comparable interest of people who called 911 and survived, and whether the injury to privacy would be "unwarranted" within the meaning of FOIL's privacy exception. Public Officers Law § 87 (2) (b), which creates the privacy exception, refers to section 89 (2), which contains a partial definition of "unwarranted invasion of personal privacy," but section 89 (2) (b) is oflittle help here; it says

27

only that "[a]n unwarranted invasion of personal privacy includes, but shall not be limited to" six specific kinds of disclosure. None of the six is relevant to this case, and so we must decide whether any invasion of privacy here is "unwarranted II by balancing the privacy interests at stake against the public interest in disclosure of the information.

In Gannett Co. v. County of Monroe, 45 N.Y.2d 954. 956 (N.Y. 1978), the Court of Appeals

said:

Specifically, we have assumed, therefore, that the county could establish by competent proof that disclosure of the items sought by Gannett would result in personal or economic hardship to the terminated county employees within the meaning of section 88 (subd 3, par e) of the Public Officers Law. Such proof, in our view. would not alter the outcome because, as noted by Justice Denman, the exception in paragraph e is available only if there is both proof of such hardships and it is established that the records sought are not relevant or essential to the ordinary work ofthe agency or municipality. The latter branch of this conjunctive requirement cannot be met in this instance.

In Buffalo News v. Buffalo Mun. Housing Authority, 163 A.d.2d 830 (4th Oep't 1990), the

Court said:

To establish this exemption, it must be shown that the disclosure would result in personal and economic hardship to the subject party and that such information is not relevant to the work of the agency maintaining it (see, Matter of Gannett Co, v, County of Monroe, 59 AD2d 309. aff'd on opinion below 45 NY 2d 9541. Here, respondent has failed to meet its burden. First, records regarding the days worked, leave taken with or without pay, and leave accrued by employees are by their very nature relevant to the day-to-day operations of the agency. Second, respondent has failed to establish how the release of these records would result

in economic or personal hardship to the subject party. The possibility that these records can be used to discover other financial information is patently insufficient. Further, respondent's claim that the newspaper may use the information to create embarrassing articles does not suffice

When in doubt in applying the personal privacy exemption, as when there are cognizable

interests on both sides of the equation, courts give priority to the public interest in disclosure. See

Matter of Rainey v. Levitt, 138 Misc.2d 962 (Sup. Ct., N.Y. Co. 1988) ("[h]ere, a liberal

construction of FOIL and a narrow interpretation of the 'unwarranted invasion of personal

28

privacy' exemption compels the conclusion that the [requested} exam scores should be provided to petitioner"). In light of these considerations, DOE made the correct determination in concluding that the TDRs are disclosable in their entirety.

In the present case, there is a strong public interest in disclosing the requested information. In being entrusted with the futures of the large majority of New York City's children, teachers are perhaps the most important public servants in the lives of New York's youngest citizens and their families. TORs provide parents with objective means to measure their effectiveness. They provide data on teacher performance in a format that may be compared to corresponding data for other teachers in the same position. As described in the Tacina Affidavit, schools are encouraged to use TORs as a tool in professional development, by using the detailed information contained therein for individual teachers as well as for all teachers in a building. Thus, for example, if a principal concludes that all 4th grade teachers are having difficulties with students in the bottom third percentile or students who are English Language Learners, school wide professional development can address this need. Individual teachers are also encouraged to use their own information to guide their own professional development. Disclosure of such records related to specific employees not only allows individual parents to know how their children's teachers are performing, but also "significantly contributes to the general public's evaluation" of the public agency itself. Powhida, supra, 147 A.D. 2d at 239.

Petitioner's argument, in contrast, may be summarized as a concern that teachers with lower ratings will be singled out and that parents will "march with their feet" out of their classrooms. This, in any event, should not negatively factor into the privacy balancing test. The TDRs, in accounting for many variables that affect student performance and in disclosing confidence intervals, provide parents with considerable information to make their own informed

29

judgments of both the effectiveness of their children's teachers and the utility of the reports

themselves. This is precisely the sort of disclosure that FOIL was intended to ensure, and the

effect of which FOIL was meant to serve ~ "greater understanding of the public in

government." See Public Officers Law § 84. The information being disclosed is drawn from

objective data analysis, and not the individual subjective analysis of a single person," The

benefit to the public of the availability of data reflecting the performance of individual teachers

without the distortion of personal observation and judgment is considerable.

Any infringement of personal privacy is therefore not, in this case, "unwarranted." Any

balancing analysis required in the application of the personal privacy exemption in this instance

mandates the disclosure of TORs.

The present case is easily distinguished from the recent Court of Appeals decision in

Matter of New York State United Teachers v, Brighter Choice Charter School et al., _N.Y.3d

_ (November 18,2010). There, petitioner, a teachers' union affiliated (like the petitioner in

the present case) with the American Federation of Teachers, requested disclosure under FOIL of

the full names of teachers at several charter schools, The schools denied disclosure under FOIL

§§87(2)(b) and 89(2)(b)(iii), which exempt from disclosure "lists of names and addresses if such

lists would be used for solicitation or fund-raising purposes" and is the sole ground in FOIL to

withhold information on the basis of the requestor's purpose. The Court upheld the denial on the

ground that the petitioner's "intent in requesting the teacher names is to expand its membership

and, by extension, obtain membership dues", In the Court's view, the request fell squarely within

the specified exemption because, as "a convenient mechanism for contacting prospective

7 We note that the final rating given by a principal is subject to disclosure, after the rating is determined final, even though the pieces that contribute to this determination are often based on subjective analysis.

30

members", it served both solicitation and fund-raising purposes. Id., Memorandum Opinion at

page 5. This is entirely different from the present case, where disclosure of TDRs with teacher

names is not being sought by a union or any membership organization; indeed, the collective

bargaining representative of New York City teachers is the petitioner in the present case, seeking

to withhold the names of teachers from disclosure. Rather, disclosure is being sought by

newspapers in order, by this means, to examine the performance of the City's schools.

In New York State United Teachers. the Court of Appeals noted that

Disclosure of the names would do nothing to further the policies of FOIL, which {arc] to assist the public in formulating "intelligent informed choices with respect to both the direction and scope of governmental activities" (Matter of Fink v. Lefkowitz, 47 N.Y.2d 567,571 [1979]. If anything, "it is precisely because no governmental purpose is served by public disclosure" of this information that section 87(2)(b )(iii)' s privacy exemption falls squarely within FOIL's statutory scheme (Matter of Federation of N.Y. Rifle & Pistol Clubs, 73 N.Y.2d at 97 [emphasis in original]). There is no indication that [New York State United Teachers] intends to use the names to, for example, expose governmental abuse or evaluate governmental activities,

Id., Memorandum Opinion at page 5. The FOIL requests in the instant case, by contrast, present

exactly those elements which the Court noted were absent in New York State United Teachers:

Because they are directed to "evaluatjing] governmental activities", they promote "the policies

of FOIL", so that a "'governmental purpose is served by public disclosure' of this information."

The balancing test applied to determine whether information may be withheld under the

personal privacy exemption points in this case clearly to disclosure.

31

CONCLUSION

Petitioner's claims that the documents at issue here are exempt from disclosure under

FOIL and that any such disclosure would violate the privacy rights of UFT members lacks merit.

Accordingly, the petition should be dismissed.

Dated:

Dated: New York, New York November 19,2010

Respectfully submitted,

MICHAEL A. CARDOZO,

Corporation Counsel of the City of New York

By: '~""~ »>:

\'" ~

/J;;SS;~~~Vi~~' v ~_~.

(. Assistant Corporation Counsel 100 Church Street, Room 2- I 06 (212) 442-3329

Steven Goulden Mark Toews

Of co unsel

32