The opinion of the court was delivered by: John M. Facciola, United States Magistrate Judge
In my initial opinion,*fn1 I permitted the search of certain backup tapes to ascertain whether that search would justify any additional searches. The search having been done, the parties could not disagree more completely as to what the search revealed. According to plaintiff, it produced useful, relevant information that justifies a second search of backup tapes for certain periods. He provides an affidavit from a forensic, computer technician who insists that a second search will not be that difficult or expensive given what the first search accomplished. Defendant disagrees and insists that the first search only produced documents that are cumulative of what plaintiff already has and that a second search would be expensive and time consuming and, therefore, completely unjustified.
A more significant consideration that has arisen is that the defendant has now ascertained that only certain backup tapes are available. Thus, the question presented has narrowed substantially to whether a search of the available backup tapes is appropriate. As my initial opinion indicates, whether a search of the backup tapes is appropriate involves an assessment of the likelihood that they will contain data (word processing documents and e-mails) that will produce information that is relevant to the lawsuit. Relevance is, in turn, a function of the relationship of the data to the two central accusations of this lawsuit: (1) that in the period from October 1994 to July 1998 the defendant retaliated against the plaintiff for plaintiff's complaints of sexual harassment against the former director of the Bureau of Prisons, Michael Quinlan ("Quinlan") and (2) that, after July 2, 1998, when plaintiff's counsel wrote a letter formally complaining about this retaliation, the defendant engaged in additional retaliatory activity that continued at least through January 12, 2002, when plaintiff's counsel informed the defendant of his intention to file this lawsuit.*fn2
Since the lawsuit claims retaliation, "relevant data" would have to mean data making it more likely than not that the true motive for the acts about which plaintiff complains was retaliatory. That data could only take only two forms.
The first would be data indicating explicit retaliation, i.e, specific references in the data to plaintiff's complaints about Quinlan.
The second would reveal the justification provided for the actions claimed by plaintiff to be retaliatory. Plaintiff would then have the right to establish that the justifications upon which the defendant relies are false, invoking the principle, familiar to Title VII jurisprudence, that the jury's disbelief of a witness's testimony permits it to draw the inference that the witness's motivation to lie is to disguise that the real reason for his actions were discriminatory or retaliatory. Waterhouse v. Williams, 292 F.2d 989, 992 (D.C. Cir. 2002)
The next phase of the analysis would be to assess the likelihood that such information exists within the period of time for which backup tapes are available. The frustration of electronic discovery as it relates to backup tapes is that backup tapes collect information indiscriminately, regardless of topic. One, therefore, cannot reasonably predict that information is likely to be on a particular tape. This is unlike the more traditional type of discovery in which one can predict that certain information would be in a particular folder because the folders in a particular file drawer are arranged alphabetically by subject matter or by author. In this case, there is the additional frustration that the tapes that do exist do not exist for clearly defined chronological periods but instead exist for certain days without any rhyme or reason for their continued existence.
Given the truly random nature of the collection of data on backup tapes, only two approaches suggest themselves. The one plaintiff suggests would be to search all the tapes insofar as they exists for the periods of time that plaintiff seeks. But, that expansion could only be justified by the possibility that there may exist data on the backup tapes that is relevant as I have defined that word, i.e., because it contains explicit references to (1) the complaints plaintiff made about Quinlan, (2) the letters his counsel wrote on July 2, 1998, or (3) his counsel's expression of an intent to file suit in a letter dated January 12, 2000, or (4) because it bears on the justification for the actions plaintiff complains were retaliatory. There is a theoretical possibility that such data exists on backup tapes but I have previously rejected the notion that the mere possibility that data exists justifies forcing the government to search backup tapes irrespective of the cost. McPeek, 202 F.R.D. at 33. Rather, the less likely it is that the backup tapes contain relevant data, the more unjust it would be to force the government to search the tapes.
The likelihood of finding relevant data has to be a function of the application of the common sense principle that people generate data referring to an event, whether e-mail or word processing documents, contemporaneous with that event, using the word "contemporaneous" as a rough guide. Conversely, it is unlikely that people, working in an office, generate data about an event that is not contemporaneous unless they have been charged with the responsibility to investigate that event or to create some form of history about it.
I will, therefore, use these principles to analyze the appropriateness of requiring additional searches of the available backup tapes.
Robert Diegelman. There are seven backup tapes available for Diegelman: January 5, 1998, January 7, 1998, January 8, 1998, January 12, 1998, April 4, 1998, April 7, 1998 and May 6, 1998. Second Declaration of Scott D. Martin, ¶ 3(a), Exhibit C to Defendant's Response to Plaintiff's Memorandum Regarding Additional Computer Discovery (hereafter "Martin"). Plaintiff made his complaint about Quinlan on June 5, 1992, and began his new job, reporting directly to Diegelman, in September, 1994. Amended Complaint, ¶¶ 20, 24. Thus, as to what I have described as "explicit retaliation", one would have to believe that Diegelman was still making references to the complaint that plaintiff had made about Quinlan six years earlier. I find the likelihood of this occurring to be so small that it cannot justify any additional search for "explicit retaliation" in the available backup tapes.
As to data bearing on events claimed to be retaliatory, the only event anywhere close in time to the dates of the available backup tapes is May 19, 1998, the date when plaintiff asked Diegelman to formally request that his position be upgraded. Amended Complaint ¶ 36. Plaintiff claims that Diegelman rejected the request because plaintiff's position was temporary and a promotion could not be justified for a position that did not have the support of the Department of Justice. Diegelman is also said to have falsely claimed that he was not the appropriate decision maker for the promotion plaintiff sought. Id.
It is unlikely, to the point of being inconceivable, that Diegelman generated data on his computer bearing on the truthfulness or legitimacy of his rejection of plaintiff's request months or even weeks before plaintiff even made the request. I, therefore, cannot see any legitimate basis on this record for any additional search of the available backup tapes.
Janis Sposato. There are eight tapes available for Sposato: June 11, 1998, March 1, 1999, April 30, 1999, June 30, 1999, July 30, 1999, August 27, 1999, November 11, 1999 and January 31, 2000. Martin, ¶ 3(d). Plaintiff seeks a search of backup tapes for Sposato's computer for the periods May 1, 1998, to August 1, 1998, and February 1, 1999 to February 1, 2000. Memorandum Regarding ...