Matter of Rayner v New York State Dept. of Corr. & Community Supervision

[*1] Matter of Rayner v New York State Dept. of Corr. & Community Supervision 2023 NY Slip Op 23293 Decided on September 14, 2023 Supreme Court, Albany County Platkin, J. Published by New York State Law Reporting Bureau pursuant to Judiciary Law § 431. This opinion is uncorrected and subject to revision before publication in the printed Official Reports.

Decided on September 14, 2023
Supreme Court, Albany County

In the Matter of the Application of Martha Rayner, Petitioner,

against

New York State Department of Corrections and Community Supervision and EQUIVANT f/k/a NORTHPOINTE, INC., Respondents.



Index No. 908549-22


Stephen Stich, Esq.
Jennifer A. Borg, Esq.
Media Freedom & Information Access Clinic, Abrams Institute
Attorneys for Petitioner
Yale Law School
PO Box 208215
New Haven, CT 06520

David A. Schulz, Esq.
Media Freedom & Information Access Clinic
Abrams Institute
Attorney for Petitioner
Yale Law School
1675 Broadway, 19th Floor
New York, NY 10019

Letitia James, Attorney General
Attorney for DOCCS
(Michael G. McCartin, of counsel)
The Capitol
Albany, NY 12224

McDowell Hetherington LLP
Attorneys for equivant
(Michael D. Matthews, Jr., of counsel)
1001 Fannin Street, Suite 2400
Houston, TX 77002

Fishkin Lucks LLP
Attorneys for equivant
(Steven M. Lucks and
Zachary W. Silverman)
500 Seventh Avenue, 8th Floor
New York, NY 10018


Richard M. Platkin, J.

Petitioner Martha Rayner brings this special proceeding under CPLR article 78, challenging the denial of a Freedom of Information Law (see Public Officers Law art 6 ["FOIL"]) request made to respondent New York State Department of Corrections and Community Supervision ("DOCCS") (see NYSCEF Doc No. 1 ["Petition"]). DOCCS and intervenor-respondent equivant, formerly known as Northpointe, Inc., oppose the Petition.

BACKGROUND

Executive Law § 259-c (4) obliges the New York State Board of Parole ("Board") to "establish written procedures for its use in making parole decisions as required by law." "Such written procedures shall incorporate risk and needs principles to measure the rehabilitation of persons appearing before the board, the likelihood of success of such persons upon release, and assist members of the state board of parole in determining which incarcerated individuals may be released to parole supervision" (id.). The Board "shall be guided by risk and needs principles, including the inmate's risk and needs scores as generated by a periodically-validated risk assessment instrument," in rendering parole decisions (9 NYCRR [*2]8002.2 [a]).

To meet these statutory and regulatory requirements, the Board has adopted a "risk assessment tool called Correctional Offender Management Profiling for Alternative Sanctions ('COMPAS') Re-entry" (Petition, ¶ 1).[FN1] COMPAS was developed by equivant (see id., ¶ 13), which licenses its proprietary risk-assessment software to criminal justice agencies throughout the United States (see id., ¶¶ 15-18). DOCCS began using COMPAS in 1998, following the award of a competitive procurement to equivant (see id., ¶ 16).

In brief, the Board uses COMPAS to generate risk and needs scores through a computerized analysis of a re-entry questionnaire (see id., ¶¶ 22-24). "For each risk, criminal involvement, or needs area, the score assigns offenders to one of ten groups, called 'deciles,' which correspond to the offender's risk level as predicted by the algorithm: low, medium, or high" (id., ¶ 25). "The range of scores COMPAS Re-entry associates with each decile — each decile's 'cut point' — is based on historical recidivism rates of a comparison class of individuals called the 'Norm Core Data Group,' divided into its own deciles based on its members' scores" (id., ¶ 26).

"This basic process is about all that's publicly known about COMPAS Re-entry in New York. The public does not know (1) how COMPAS Re-entry combines the questionnaire answers to generate the risk and needs scores, (2) each decile's cut point (i.e., what score is required to be assigned to each decile), (3) how deciles are assigned to the qualitative 'low,' 'medium,' and 'high' risk levels, and (4) the Norm Core Data Group used to generate the scores associated with each decile" (id., ¶ 28).

And "there are no published studies validating the predictive accuracy of COMPAS Re-entry in New York" (id., ¶ 29). "Northpointe and state agencies have published validation studies of other COMPAS instruments. They have also published records, for other COMPAS instruments, indicating how deciles are assigned to qualitative risk levels" (id., ¶ 30).

Petitioner, a clinical law professor at Fordham Law School, requested information from DOCCS about COMPAS to obtain "a meaningful explanation of the basis for the state's decisions to grant or withhold parole" (id., ¶ 4). Specifically, on January 4, 2022, petitioner requested the following five categories of DOCCS records under FOIL:

1. Any and all documents providing the formulas for calculating the risk [*3]and needs scales used in New York's COMPAS instrument. Including:a. Risk Scores: the variable inputs, transformations (if any) of those variables, and the coefficients applied to those variables for the following: i. Risk of Felony Violence: ii. Arrest Risk: iii. Abscond Riskb. Criminogenic Needs Scales: the question inputs, how to score each question input (i.e., what number value is assigned to which answers), how they are combined to yield the scale score for the following: i. Criminal Involvement ii. History of Violence iii. Prison Misconduct iv. Reentry Financial v. Reentry Employment Expectations vi. Low Family Support vii. Negative Social Cognitions viii. Low Self-Efficacy/Optimism ix. Reentry Substance Abuse2. All documents indicating the decile cut-points for each: a. Risk Score (listed in 1.a. (i) — (iii) above); b. Criminogenic Needs Scales (listed in 1.b. (i) — (xii) above).3. All documents indicating how deciles are assigned to the corresponding qualitative values (e.g., "Low, Medium, High") for each Risk and Needs Scale.4. The Norm Core data group against which New York is scoring individual COMPAS assessments.5. All validation studies of NYCOMPAS instrument (NYSCEF Doc No. 2, Ex. A).

DOCCS denied the FOIL request in its entirety on June 13, 2022 (see id., Ex. K). Initially, DOCCS determined that petitioner's "request for any and all records pertaining to the formulas for calculating the risk and needs scales used in New York's COMPAS instrument is overly broad . . . [and] fails to reasonably describe records as required by Public Officers Law § 89 (3) (a)" (id.). And insofar as DOCCS was able to identify records responsive to petitioner's request, the records were determined to be exempt from disclosure as trade secrets or material that would cause substantial competitive harm to equivant if disclosed (see id.).

Petitioner's administrative appeal was denied on July 21, 2022 for essentially the same reasons (see NYSCEF Doc No. 2, Ex. M).[FN2]

Petitioner commenced this CPLR article 78 proceeding against DOCCS on November 15, 2022, challenging the denial of her FOIL request (see NYSCEF Doc Nos. 1-7).

In its answering papers in opposition to the Petition, DOCCS relied on the affidavit of Eugenie Jackson, PhD, the statistician who leads equivant's research into risk and needs assessments (see NYSCEF Doc No. 14 ["Dr. Jackson Aff."], ¶¶ 3-6). Dr. Jackson attests that the proprietary formulae, algorithms, risk and needs scales, and norming data incorporated into New York's customized implementation of COMPAS Re-entry ("COMPAS-NY") are trade secrets that belong to equivant (see id., ¶ 10).

Prior to the return date, equivant moved for leave to intervene to protect its trade secrets (see NYSCEF Doc Nos. 18-20). Petitioner did not oppose intervention, but filed a lengthy memorandum taking the position that the technology underlying COMPAS-NY is not entitled to trade secret protection (see NYSCEF Doc No. 21). In reply, equivant presented additional arguments and evidence in support of its claim of trade secrecy (see NYSCEF Doc Nos. 22-25). By order dated April 27, 2023, the Court ordered equivant joined as a respondent (see NYSCEF Doc No. 32).[FN3]

Oral argument was held on August 24, 2023, DOCCS provided the Court with responsive records for in camera inspection on August 31, 2023, and a copy of the certified transcript of oral argument was filed on September 13, 2023 (see NYSCEF Doc No. 44). This Decision, Order & Judgment follows.

DISCUSSION

A. Legal Framework

Under FOIL, agency records presumptively are available for public inspection unless they fall within a specific exemption set forth in Public Officers Law ("POL") § 87 (2) (see Matter of Data Tree, LLC v Romaine, 9 NY3d 454, 462 [2007]; Matter of Aurigemma v New York State Dept. of Taxation & Fin., 128 AD3d 1235, 1236-1237 [3d Dept 2015]). Respondents "bear[] the burden of demonstrating that the requested material falls squarely within an exemption by articulating a particularized and specific justification for denying access" (Matter of MacKenzie v Seiden, 106 AD3d 1140, 1141 [3d Dept 2013] [internal quotation marks and citations omitted]; accord Aurigemma, 128 AD3d at 1237; see Matter of [*4]Rose v Albany County Dist. Attorney's Off., 111 AD3d 1123, 1125 [3d Dept 2013]).

At issue in this proceeding are the exemptions of POL § 87 (2) (d), which authorize the withholding of records that "are trade secrets or are submitted to an agency by a commercial enterprise or derived from information obtained from a commercial enterprise and which if disclosed would cause substantial injury to the competitive position of the subject enterprise." POL § 87 (2) (d) encompasses at least two distinct types of confidential commercial information imparted to government agencies.

First, agency records are exempt from disclosure if they "contain bona fide trade secrets" (Matter of Verizon NY, Inc. v New York State Pub. Serv. Commn., 137 AD3d 66, 74 [3d Dept 2016], affg 44 Misc 3d 858 [Sup Ct, Albany County 2014, Ferreira, J.]). Although the term "trade secret" is not defined in FOIL (or elsewhere in New York's statutory law), courts generally follow Section 757 of the Restatement of Torts in determining whether information is entitled to protection (see Ashland Mgt. v Janien, 82 NY2d 395, 407 [1993]; Verizon, 137 AD3d at 72-73).

Under the Restatement, a trade secret is "'any formula, pattern, device or compilation of information which is used in one's business, and which gives [the business] an opportunity to obtain an advantage over competitors who do not know or use it'" (Ashland, 82 NY2d at 407, quoting Restatement of Torts § 757, Comment b). The factors to be considered in determining whether given information is one's trade secret are:

(1) the extent to which the information is known outside of [the] business; (2) the extent to which it is known by employees and others involved in [the] business; (3) the extent of measures taken by [the business] to guard the secrecy of the information; (4) the value of the information to [the business] and [its] competitors; (5) the amount of effort or money expended by [the business] in developing the information; (6) the ease or difficulty with which the information could be properly acquired or duplicated by others (id.).

Second, even records that do not qualify for trade secret protection are exempt if "submitted to the agency by a commercial enterprise," and disclosure "would cause substantial injury to the competitive position of the subject enterprise" (POL § 87 [2] [d]; see Verizon, 137 AD3d at 69). "[T]he party seeking [this] exemption must present specific, persuasive evidence that disclosure will cause it to suffer a competitive injury" (Matter of Markowitz v Serio, 11 NY3d 43, 51 [2008]).

[*5]B. Analysis

Respondents maintain that that the formulae, algorithms, risk and needs scales, and norming data underlying COMPAS-NY are exempt from disclosure under POL § 87 (2) (d) as bona fide trade secrets.

In her affidavit, Dr. Jackson explains that COMPAS-NY assesses risk and needs associated with re-entry through application of the formulae, algorithms, risk and needs scales, and norming data incorporated into its flagship COMPAS product, as customized for DOCCS to reflect the specific characteristics of New York's offender population (see Dr. Jackson Aff., ¶¶ 13-14).[FN4]

As the leader of equivant's research department, Dr. Jackson attests that COMPAS and the New York-specific refinements were developed by equivant through substantial effort (see id., ¶ 15; see also id., ¶ 25).

With regard to secrecy, COMPAS-NY and its "underlying data are generally unknown outside of equivant's business" and outside of DOCCS's, which is the "only customer" for COMPAS-NY (id., ¶ 19). In fact, the particular customizations in COMPAS-NY "are known only by the current and former equivant employees that developed them and [by DOCCS personnel]" (id., ¶ 20).

Additionally, equivant protects the secrecy of COMPAS by restricting access to its technology, both inside and outside of the company (see id., ¶¶ 19-21), opposing subpoenas directed at the production of COMPAS materials (see id., ¶ 22), and opposing the production of COMPAS materials in lawsuits against customers (as equivant did here) (see id., ¶ 23).[FN5]

Finally, Dr. Jackson attests that COMPAS-NY is the product of substantial investments by equivant, beginning with its research and development of COMPAS, and then the design and implementation of the New York-specific refinements (see id., ¶ 24), which was "a tremendous effort by numerous employees that took countless hours" (id., ¶ 25; see also NYSCEF Doc No. 23). And given the magnitude of equivant's investments, it would be very difficult for a competitor to properly duplicate the functionality of COMPAS-NY (see Dr. [*6]Jackson Aff., ¶¶ 26-28).

In arguing that the foregoing evidence is insufficient to discharge respondents' burden of demonstrating that the formulae, algorithms, risk and needs scales, and norming data underlying COMPAS-NY are trade secrets, petitioner argues principally that the record of this proceeding "lacks the 'specific, persuasive evidence' of competitive harm that the law requires" (NYSCEF Doc No. 21 at 3-4 [internal quotation marks and citation omitted]).

Petitioner acknowledges, as she must, that respondents' invocation of the FOIL exemption for trade secrets obviates any need for a showing that disclosure of COMPAS-NY "would cause substantial injury to the competitive position of [equivant]" under POL § 87 (2) (d). Nonetheless, petitioner insists that essentially the same showing is required under general principles of trade secret law.

In making this argument, petitioner relies on the Restatement factor calling for consideration of "the value of the information to the business and its competitors" (Verizon, 137 AD3d at 73 [internal quotation marks and citation omitted]). According to petitioner, "FOIL's trade-secret exemption requires proof of actual competition and competitive harm because New York's definition of a 'trade secret' requires such proof" (NYSCEF Doc No. 21 at 4).

The Court is unpersuaded by petitioner's argument, which represents a misapplication of the law of trade secrets. Dr. Jackson's affidavit provides particularized and specific testimony that the formulae, algorithms, risk and needs scales, and norming data underlying COMPAS-NY — including the manner in which risk and needs scores are derived from the Board's questionnaire, the decile cut-off points used by COMPAS-NY, and the manner in which deciles are assigned to qualitative categories (see Petition, ¶ 28) — were developed, identified, gathered and/or analyzed by equivant, "a private, for-profit corporation" (NYSCEF Doc No. 5 ["MOL"] at 1), through substantial investments of time, money and effort.

To recoup these investments and continue its research and development into risk and needs assessment instruments, equivant licenses software incorporating its proprietary technology to DOCCS (see NYSCEF Doc No. 4, Ex. A ["Contract"]).

But if petitioner's FOIL application were granted and equivant's technology opened to the public, it would be a simple matter for one of the State's information technology vendors to supply DOCCS with the functionality of COMPAS-NY at a much lower cost. After all, a software developer given access to the inner workings of COMPAS-NY would not have to invest "substantial money and effort" in developing the "scales, formulas, and analysis" used within the program to generate scores and qualitative assessments (Dr. Jackson Aff., ¶ 26).

Thus, continued secrecy of COMPAS-NY has real value to equivant and gives it a clear "advantage over competitors" that do not have access to the scales, norming data, and cut-off points incorporated into the program (Restatement of Torts § 757, Comment b; see also Restatement [Third] of Unfair Competition § 39 ["potential economic advantage over others"]; id., Comment e ["advantage" need only be "more than trivial"]).

Moreover, petitioner's argument rests on the false premise that equivant "faces no competition with respect to the New York COMPAS Re-entry tool. The [Board] is, essentially, statutorily required to use the product, it is customized for New York, and [equivant] has not faced a competitive bid in New York since the 1990s" (MOL at 19; see Petition, ¶¶ 16-18).

New York law requires the Board to use risk and needs assessments in making parole decisions (see Executive Law § 259-c [4]; 9 NYCRR 8002.2 [a]), but there is no requirement that equivant's COMPAS-NY product be used for that purpose. And if other vendors could replicate the functionality of COMPAS-NY using equivant's formulae, algorithms, risk and needs scales, and norming data at a far lower cost, it seems doubtful that DOCCS would be allowed to continue with non-competitive procurements (see State Finance Law § 163 [1] [h] [requiring agency to document "the basis upon which it determined the cost was reasonable"]).

The Court therefore concludes that the value of the risk and needs assessment technology embodied in COMPAS-NY, both to equivant and to potential competitors, "is self-evident and its disclosure would almost assuredly complicate, if not compromise, [equivant's] competitive position" in New York (and perhaps in other states) (Matter of City of Schenectady v O'Keeffe, 50 AD3d 1384, 1386 [3d Dept 2008], lv denied 11 NY3d 702 [2008]).

Petitioner further argues that respondents' claim of trade secrecy is undermined by the fact that equivant and criminal justice agencies "have published information of the type that [DOCCS] has withheld" (MOL at 17). "For example, [equivant] and [New York State] have published documents indicating which deciles are assigned to which qualitative values for COMPAS Re-entry in other jurisdictions and another COMPAS tool used in New York, COMPAS-Probation" (id. at 7; see NYSCEF Doc No. 4, Ex. E).

But petitioner makes no claim that respondents have disclosed the inner workings of COMPAS-NY or the manner in which the program converts a completed parole re-entry questionnaire into risk and needs scores with qualitative assessments. General disclosures about COMPAS-NY or detailed disclosures about other implementations of COMPAS (see NYSCEF Doc No. 4) do not [*7]disclose the inner-workings of COMPAS-NY (see MOL at 5 ["the public knows only the outlines of COMPAS Re-entry in New York."]).[FN6] There simply has been no showing that the material respondents seek to protect is known by others.

Petitioner also contends that equivant waived its claim of trade secrecy by failing to designate documents as confidential under its contract with DOCCS (see Contract at 25-26). The Court disagrees. Even assuming that equivant's alleged failure to designate amounted to a waiver of its right to insist that DOCCS invoke FOIL's trade secret exemption (see e.g. Matter of Lockheed Martin IMS Corp. v New York State Dept. of Family Assistance, 256 AD2d 847, 849 [3d Dept 1998]), DOCCS joins with equivant in arguing that COMPAS-NY's technology is a trade secret, and petitioner has no standing to assert waiver on DOCCS's behalf.

Based on the foregoing, the Court concludes that the technology underlying COMPAS-NY is a bona fide trade secret belonging to equivant, and agency records containing such material are exempt from disclosure under POL § 87 (2) (d).[FN7] Any change to this rule based on the "uniquely important public interest" of parole oversight (MOL at 20-21) must come from the State Legislature.

Finally, DOCCS denied the first prong of petitioner's FOIL request, seeking "[a]ny and all documents providing the formulas for calculating the risk and needs scales used in New York's COMPAS instrument," as overly broad and insufficiently descriptive. Insofar as the request is directed at the risk and needs scales used in COMPAS-NY, it calls for the disclosure of exempt trade secrets. And to the extent that petitioner is seeking "any and all" documents concerning COMPAS-NY's risk and needs scales, the Court agrees with DOCCS that the request is overly broad and insufficiently particularized.

[*8]CONCLUSION

Accordingly, it is

ORDERED and ADJUDGED that the Petition is dismissed.

This constitutes the Decision, Order & Judgment of the Court, the original of which is being uploaded to NYSCEF for entry by the Albany County Clerk. Upon such entry, counsel for DOCCS shall promptly serve notice of entry on all parties entitled to such notice.

Albany, New York
September 14, 2023
RICHARD M. PLATKIN
A.J.S.C.

Papers Considered:

NYSCEF Doc Nos. 1-5, 12-15, 17, 19, 21-25, 44 and in camera materials. Footnotes

Footnote 1: Except as may otherwise be indicated, references to COMPAS hereinafter shall refer to the COMPAS Re-entry functionality.

Footnote 2: The initial FOIL denial also claimed that responsive records were exempt from disclosure under Public Officers Law § 87 (2) (a) by reason of Section 106 of the Copyright Act of 1976, but the appeals decision abandoned that rationale.

Footnote 3: The papers filed by equivant in support of its motion to intervene were deemed an answer in opposition to the Petition and in support of DOCCS's answer, petitioner's opposition to equivant's arguments was deemed a reply to equivant's answer, and equivant's reply to petitioner's opposition was deemed an authorized sur-reply (see id.).

Footnote 4: The State's contract with equivant makes any New York-specific improvements and customizations to COMPAS the property of equivant (see NYSCEF Doc No. 4, Ex. A [B] [3], [6]).

Footnote 5: See e.g. Flores v Stanford (2021 WL 4441614, *1, 7 n 7; 2021 US Dist LEXIS 185700, *5-6, 21 n 7 [SD NY, Sep. 28, 2021, No. 18 Civ. 02468 (VB) (JCM)]); State v Loomis (371 Wis 2d 235, 258, 881 NW2d 749, 761 [2016], cert denied — US —, 137 S Ct 2290 [2017]).

Footnote 6: Petitioner concedes as much in contending that knowledge of the particulars of COMPAS-NY would not assist competitors in other jurisdictions (see MOL at 16-17; see also id. at 4, 16 [observing that different states' implementations of COMPAS "take different inputs" and "each state's outputs are customized"]).

Footnote 7: The Court has reviewed the 326 pages of responsive records submitted by DOCCS for in camera inspection, each page bearing the legend "Highly Confidential," and finds that the records properly were withheld as bona fide trade secrets. To the extent that the Court can discern the presence of any non-trade secret information (i.e. publicly available information about COMPAS) within the records, it is so inextricably intertwined with exempt material that redaction would be impracticable and result in the production of heavily-redacted records that disclose no additional information about the inner workings of COMPAS-NY.