Ethics of (Not) De-duping (from LinkedIn’s e-Legal)
Once again, Anne Kershaw and team (here, Joe Howie) have done a wonderful job documenting for potential judicial review, practices that have become standard. With the current climate spotlighting potential cost reductions, this is certainly a ripe area. Joe’s summary on the article:
Survey shows half of projects don’t use across-custodian deduping causing 20% higher review volumes and higher costs; duplicative review is unnecessary and hence unethical.
A Law Technology News article by Anne Kershaw and me summarized the results of a survey we conducted of leading e-discovery providers on the topic of deduping: 21% average volume reduction achieved by deduping within single custodians, 38% average volume reduction achieved by deduping across custodians. The article also analyzed the results in terms of ethics: duplicative reviews are unnecessary and hence unethical. See http://tinyurl.com/dedupingethics . I’d be curious to get feedback on this. Is there an ethical obligation to be efficient or at least reasonably efficient? Are there some justifications for single-custodian deduping we don’t address in the article or the report?
The authors claim it is unethical to review duplicates because it is unnecessary and adds costs. This is an area where a client may have a preference, as one law firm may be handling part of a case where another is handling another part. There would need to be a system where the law firms could have transparency into what the other had collected and produced.
If there is one law firm or the multiple firms can collaborate, I would agree that reviewing dups for the first pass, certainly, is unnecessary and would drive up costs.
Once the witness preparation starts, I would argue that it is unethical not to review the relevant material for each witness in the context of their contribution to the instant case, a small subset of the original data.
It is important for the client to define “de-dup” when working with a service provider. For example, de-duping at point of collection or prior to publishing can eliminate the possibility of knowing another custodian held the same material unless there is sufficient logging in a database. In many cases, even when logging occurs, the review strategy does not take them into account.
Most providers have standardized on de-duping emails at the “chain” level rather than the “item” level. This is to avoid review and production of pieces of a document, sometimes called producing Swiss cheese. This production might have a couple attachments, but no email body or vice versa. Many “cases” are actually multiple matters, with productions to multiple requestors. It is important to be able to parse all of these levels of de-duplication so that the process does not assume that a production to one requestor is enough to suffice for a different requestors.
Once the privilege review starts, the decisions made early for de-duplication can cause ethical issues. For example, de-duping at the item level (an PDF attachment can be a duplicate of a PDF in a “my documents” folder) can cause issues when that same .pdf can be an attorney-client communication in the first context and work product in the second.
The takeaway? Unless it is a very simple case, de-dup liberally but strategically, and set it up to accommodate the most complex situations, such as:
- Multiple law firms in the same matter
- Multiple requestors in the same matter
- The ability to generate reports and productions that include the duplicates (and their contexts–custodian, media and path) skipped in publishing for review or the review itself
- De-dup at the chain level for review and production
- Identify duplicates at the item level so that privilege review can be expedited by viewing individual files in multiple contexts
Filed under Sound Evidence, Technology Counsel.