The opinion of the court was delivered by: Lewis A. Kaplan, United States District Judge
This case again is before the Court on yet another discovery problem.
In a consent order dated December 3, 2002, this Court directed the City to produce a computer disk or disks containing the Police Department's CATs database, updated for all cases that were in the system as of March 2000. The database reflects, perhaps among other things, the charges, dispositions and penalties imposed in departmental disciplinary cases and thus may be important in proving or refuting plaintiffs' claims of a pattern or practice of racial and ethnic discrimination in the disciplinary system.
A disk was produced in December 2002. Plaintiffs thereupon noticed some problems with the data, which they brought to defendants' attention. On January 30-31, 2003, defendants turned over what they believed was a final, properly formatted disk. Plaintiffs' expert thereupon prepared what was believed to be his final report.
In late February, the City discovered that the data on the late January disk was not accurate. Plaintiffs sought the Court's assistance — they asked that it direct that all expert reports be based on the allegedly inaccurate late January data and that defendants be precluded from challenging the accuracy of that data. The Court denied that application by endorsed order on February 28, 2003. It directed, however, that the defendants "identify in what way, if any, the data and Plaintiffs' findings are affected by the latest claimed problem with the database." Endorsed order, Feb. 28, 2003.
On March 25, 2003, the City submitted a declaration of Dr. Ali I. Saad, who holds a Ph.D. in economics and is a managing partner of a firm that conducts economical and statistical analyses. The declaration details the differences between the original CATs data produced by defendants in this action ("Old CATS") and that provided in late January 2003 ("New CATS"). Without recounting the comparison in detail, Dr. Saad makes the following important points:
• The Old CATS "data most accurately reflects the
actual case facts, in terms of what the officer was
charged with, what the dispositions of each charge
were, and the penalties the officer received." Saad
Decl. ¶ 4.
• The New CATS disk "does not accurately reflect
the information contained in the actual paper case
files." Id. ¶ 5.
• The Old CATS data reflects 4,698 cases against
officers of all races and the dispositions and
outcomes of about 3,600 of them. While the New CATS
data was intended to update the roughly 1,000 cases
open at the time the Old CATS data was compiled, the
attempt to update failed at least in the respect
that the migration of data on the closed cases from
Old CATS to New CATS resulted in at least some
unintended changes in the recorded dispositions.
Id. ¶ 6. A review of a sample of paper logs
revealed that in each case in which a discrepancy
appeared, the Old CATS data was correct and the New
CATS data incorrect. Id. ¶¶ 7-8.
• The New CATS data also do not in all cases
correspond to the paper records with respect to
updated data — in other words, Dr. Saad found
instances in which the information concerning the
disposition of cases that were open at the time the
Old CATS data was compiled does not correspond to
the paper records concerning the disposition of
those cases. Id. ¶¶ 9-10.
• The New CATS database contains duplicate
records, which would affect the accuracy of any
statistical analysis. Id. ¶¶ 12-14.
• As there is no evidence that the resolutions
for the roughly 1,000 cases that were open at the
time the Old CATS data were compiled differ
systematically from the resolutions of the 3,600
cases for which Old CATS concededly contains
accurate information, the Old CATS data is entirely
"sufficient to establish whether or not any
statistical patterns exist." Id. ¶ 17.
Plaintiffs have not challenged the accuracy of Dr. Saad's account. It therefore is perfectly apparent that resting any statistical analysis of departmental discipline on the New CATS data would yield a product of doubtful admissibility and even more doubtful utility. Nor have plaintiffs challenged defendants' explanation of the computer system changes and data migration problems that caused the problem. Plaintiffs nevertheless contend that defendants have violated both the Court's December 3, 2002 order, which required production of an updated CATS database, and the February 28, 2003 order insofar as they have failed, allegedly, "to identify in what way, if any, . . . Plaintiff's findings [based on the New CATS data] are affected by the latest claimed problem with the database." Letter, Richard A. Levy, Esq., to Court, Apr. 2, 2003. They seek sanctions consisting of (a) denial of defendants' motion for summary judgment on the merits, (b) an adverse inference instruction at trial based on "loss or corruption of the data," and (c) preclusion of any challenge by defendants to plaintiffs' statistical report based on the New CATS data. In addition, the seek an order directing defendants to correct any flaws in the database by April 15, 2003 and to pay plaintiffs' costs and fees in producing a new statistical report. Defendants, for their part, deny any impropriety and have undertaken to correct the data problem by May 12, 2003.
Plaintiffs' frustration is understandable. This application is not.
First, it is undisputed that the New CATS data is inaccurate and, in consequence, that any conclusions that might be drawn therefrom would be worthless. There is no responsible case for insisting that the case be decided on the basis of evidence that everyone agrees is wrong. That is why the Court denied plaintiffs' prior application to preclude defendants from challenging an expert report based on erroneous data. That is why it denies this one.
Second, there is not the slightest hint of culpable behavior on the part of the defendants. To suggest the imposition of such harsh sanctions in this context is not helpful.
Third, the argument that defendants are in violation of the direction that they "identify in what way, if any, . . . Plaintiff's findings [based on the New CATS data] are affected by the latest claimed problem with the database" in the face of the Saad declaration is just silly. The Saad declaration is crystal clear. Paraphrased in simple terms, it says that the New CATS data is garbage and that plaintiffs' expert — through no fault of his or of plaintiffs — based his analysis on what we now know is garbage. "Garbage in, garbage out."
It is extremely doubtful, in light of paragraph 17 of Dr. Saad's declaration, that updating the Old CATS database is an exercise worth the time and effort. Certainly plaintiffs' expert has not attempted to refute his view. Nevertheless, inasmuch as defendants consented to an order requiring them to do so, however, the Court will hold them to their bargain. In consequence, plaintiffs' application is denied in all respects save ...