tl;dr: Recent appeals court ruling means (1) lawyers doing document review may be eligible for overtime back-pay and (2) machines might not be capable of UPL
On July 23, 2015, the Court of Appeals for the Second Circuit released a decision in which it determined that contract attorneys performing document review may not be engaging in the “practice of law”, at least in North Carolina. This could create a responsibility to pay millions of dollars in unpaid overtime, including by some of the largest law firms in the world (in this case, Skadden) and an entire industry of legal staffing companies that provided lawyers to law firms on a contract basis. That said, the court did not articulate a clear standard for the circumstances in which document review does not constitute the practice of law.
In addition, perhaps unintentionally, the definition the court used to describe work devoid of legal judgment – work that “otherwise could be performed entirely by a machine” – has implications for unauthorized practice of law by machines. On its face, the ruling suggests that a machine can never engage in the unauthorized practice of law.
Potentially disastrous consequences for law firms and legal staffing companies
The question directly before the court in Lola v Skadden was whether Lola, a contract lawyer doing document review in North Carolina, can benefit from overtime protections afforded to non-professional employees. Lawyers with a “valid license or certificate permitting the practice of law” who are “actually engaged in the practice thereof” are not entitled to overtime pay, a carve-out that is highly beneficial to employers operating in an industry notorious for long hours. As the plaintiffs all held valid law licenses (although not necessarily in North Carolina, a fact that should have been addressed in more detail), the relevant issue concerned the practice of law. In declining to dismiss the lawsuit, the court accepted the possibility that some document review can be so routine that it is not, in fact, legal work.
By ruling that standard labour protections might apply to lawyers engaged in routine work, the court potentially exposed many large law firms and legal staffing companies to enormous back-pay liabilities. The decision could open the door to copycat class actions by contract lawyers performing comparable work, similar to how the decision awarding minimum wage to two unpaid Fox Searchlight interns for the 2010 movie Black Swan (recently overturned on appeal) encouraged lawsuits against NBCUniversal, ICM Partners, Viacom and Lionsgate.
These potential consequences come with a number of caveats, of course. First, Lola is a preliminary decision without any findings of fact, and it does not clearly articulate the circumstances in which document review is no longer the practice of law, other than recognizing that this possibility exists. Second, it is disputed whether law firms are in an employer relationship with respect to the contract lawyers, whether the legal staffing companies alone will be responsible for paying damages, and whether the underlying clients will pay either way. Third, anecdotally, a number of legal outsourcers and law firms pay overtime for contract lawyers, so would not face any additional liability. Fourth, the definition of “practice of law” is state-specific (discussed in more detail below).
The money tied up in this non-payment-of-overtime issue could be significant. Document review accounts for a huge percentage of the overall legal spend on litigation and various types of regulatory disclosure. On large, complicated matters, the cost of document review can be millions of dollars. A recent study of eDiscovery costs for large corporations found that median eDiscovery costs on each matter were $1.8 million, of which the largest cost component was the work at issue in Lola - review of documents for relevance, responsiveness, and privilege - which accounted for about 73 percent of total eDiscovery costs.
According to the figures in Lola, contract lawyers conducting document review could stand to receive back-pay for in the neighbourhood of 11%-27% of total compensation. The plaintiff in Lola claimed to have worked roughly 45-55 hours a week without receiving time-and-a-half compensation for time worked in excess of the standard workweek in North Carolina, 40 hours.
In addition, the collection rate on any settlement for this type of case would be very high. On any class action, there is a portion of the plaintiff class that cannot be contracted or choose not to collect damages. For example, in a class action on behalf of the victims of toxic contamination - a la the plot of A Civil Action starring John Travolta - the permanent residents of the affected area might be identifiable, but the plaintiffs who passed through the town temporarily might be more difficult to contact and it might be difficult to ascertain any illness they suffered from minimal exposure. That would not be the case here: employers would have contact information for all plaintiffs and the damages owed to each lawyer would be clear on the basis of timesheets and payment records.
Obviously, estimating the total cost at stake in Lola is difficult because document review expenses are rarely publicly disclosed and many important issues remain unresolved. Suffice to say, however, that the combined damages could be enormous, particularly if Lola is followed in other states. There are good reasons to believe this will be the case.
With this definition of “practice of law”, can a machine ever commit UPL?
The court in Lola decided that “inherent in the definition of ‘practice of law’ in North Carolina is the exercise of at least a modicum of independent legal judgment”. Without legal judgment, in other words, there is no “practice of law”. Importantly, the court stated that tasks that could be performed entirely by machines cannot, by definition, involve legal judgment – and therefore do not constitute the practice of law:
A fair reading of the complaint in the light most favorable to Lola is that he provided services that a machine could have provided. The parties themselves agreed at oral argument that an individual who, in the course of reviewing discovery documents, undertakes tasks that could otherwise be performed entirely by a machine cannot be said to engage in the practice of law.
As noted insightfully in this post, this might be a misguided standard. On the one hand, work without any legal judgment is an extremely low bar; many secretaries and law clerks perform work under the supervision of lawyers that easily exceeds this standard. On the other hand, lots of currently available legal software is capable of replicating quite sophisticated legal judgment. Our software at Kira, for example, helps lawyers find, extract, and analyze complicated contractual provisions like change of control or assignment with greater accuracy than similar lawyers not using it.
Regardless of its internal (in)consistency, this standard has direct implications for unauthorized practice of law by machines, at least in North Carolina. If a machine cannot, by definition, engage in the practice of law, how can software ever be responsible for unauthorized practice, even when legal software is given directly to end users without mediation or oversight by a lawyer?
The “machines cannot practice law” standard is a welcome contradiction to some existing case law. As an illustration, consider LegalZoom, a legal technology company that offers software to assist with the preparation of legal documents. LegalZoom is currently the focus of a lawsuit for unauthorized practice of law in North Carolina. The North Carolina Bar Association, which initiated the proceedings against LegalZoom, contends that the branching process used by LegalZoom to identify relevant customer information is akin to the manner in which a lawyer interviews a client and requires professional judgment.
The fact that LegalZoom’s templates were completed by a machine using a consistent and predictable process was insufficient to dismiss the case: the Superior Court of North Carolina postponed judgment until a more complete factual record is developed. Although LegalZoom has been more successful in other states in which it has been pursued for unauthorized practice of law, the North Carolina decision is evidence that there is not, as yet, a national consensus.
Applicability of this decision to other jurisdictions
The extent to which Lola will apply outside of North Carolina remains to be seen.
State, not federal, law informs the definition of “practice of law”, so the holding of the case is only directly applicable to North Carolina. As a Second Circuit decision, Lola is binding precedent as to North Carolina law for federal courts in New York, Connecticut, and Vermont. Who cares about the interpretation of North Carolina law by New York courts? Well, most state definitions of “practice of law”are similar to North Carolina’s and, at the least, there is a fair amount of document review done in Second Circuit jurisdiction New York. Further, the Second Circuit is a very influential court so the reasoning of the decision may be adopted by other courts interpreting similar definitions of “practice of law”.
New York State
Since Lola involves the interpretation of North Carolina’s legislation, a New York court could, in theory, reach a different conclusion about the meaning of “practice of law” within New York, even though Lola is binding precedent for Second Circuit courts.
Nothing to date about the treatment of “practice of law” in New York suggests the definition in New York is substantially different from North Carolina, so Lola will likely be followed closely in New York. New York case law is also broadly consistent with North Carolina law on this issue. In particular, the court in Lola cited a leading New York case about the practice of law necessarily involving bespoke legal advice and opinions: In re Rowe, 80 N.Y.2d 336, 341-42 (1992). Rowe held that authoring an article on the legal rights of psychiatric patients who refuse treatment did not constitute the practice of law because “[t]he practice of law involves the rendering of legal advice and opinions directed to particular clients”.
Application outside of NC and Second Circuit States
Lola will be persuasive authority elsewhere in the country, and, barring precedent from the Supreme Court or another appellate court, should be followed closely since state definitions of “practice of law” are relatively consistent.
Several years ago, the ABA compiled a list of state definitions of “practice of law”. One shared trait is that the definitions are vague regarding whether machines can engage in the practice of law. Consider, for example, the leading case in California, People v. Merchants Protective Corp., 209 P.363, 365 (1922) (quoting a case from the Court of Appeals for Indiana):
As the term is generally understood, the practice of the law is the doing or performing services in a court of justice, in any matter depending therein, throughout its various stages, and in conformity to the adopted rules of procedure. But in a larger sense it includes legal advice and counsel, and the preparation of legal instruments and contracts by which legal rights are secured although such matter may or may not be depending in a court.
Even extremely detailed definitions gloss over whether they apply to machines. In DC, for example, rule 49(b)(2) of the Court Rules of the DC Court of Appeals includes four separate methods of definition, none of which distinguish between the lawyers who train and create software to solve legal problems and the software itself.
Another common thread across state definitions is the emphasis on legal judgment, which was highlighted in Lola. Sometimes this is expressed as a distinction between legal information and legal advice or representation, where legal advice or representation is tailored to the specific client’s situation. To the extent another court is looking to define “legal judgment” as part of a broader discussion of about the practice of law, Lola will be relevant regardless of the state jurisdiction.
Next Steps for Lola...
The case was remanded to a North Carolina court for consideration of evidence to determine whether Lola (and the similarly situated plaintiffs employed by Tower) were, in fact, engaging in the practice of law and, if so, who should be responsible for any outstanding overtime pay. The Second Circuit decision merely accepted that this outcome was a possibility, if all of the facts were as pleaded by the plaintiffs. Irrespective of how this turns out, the decision in Lola should have real impacts on document reviewers past and future, as well as automators of legal work.