The dramatic increase in electronic information during the past decade, combined with the decline of legal budgets and the rise in hourly rates, have prompted lawyers, clients and information-retrieval experts to look for alternatives to the traditional linear method of document review in civil litigation. The allure of an all-technology answer to the problem of managing large volumes of documents is understandable, given society’s infatuation with electronics. Computer chips enable smartphones to find nearby restaurants, so it seems reasonable to assume that a computer should be able to do something as simple as identify a relevant document in a group of emails.
Whether document reviews will be entirely computerized in the future remains to be seen. The issue before the legal community today is whether predictive coding, also called computer-assisted or technology-assisted review, can produce quality document reviews that satisfy clients’ needs and meet attorneys’ professional obligations. Two studies are frequently cited for the proposition that predictive coding is just as good as, if not better than, manual review. Although both suggest the need for future research in this area, neither one provides definitive evidence of the superiority of the computer over a properly conducted manual review.
