10290820743_f98f5289b5_z

The Data Revolution in Law Started Decades Ago

Editor’s Note: The author of this post, Daniel Lewis, is the CEO of Ravel Law, a legal analytics company he co-founded in 2012 while attending Stanford Law School. 

 

By Daniel Lewis, Co-Founder and Chief Executive Officer, Ravel Law

The transformative impact of data science and analytics in fields like sports and politics is well known, and every day there seems to be another “Moneyball for X” analogy. But what, if anything, does this mean for the legal world, and when will it happen? This is a story about the data revolution that is already transforming the law and how its potential was foretold long ago.

In 1948, years before he became president, Lyndon B. Johnson fought a desperate senate election campaign. A loss would ruin his political career. With election day approaching, a district court judge issued an injunction keeping Johnson’s name off the ballot until accusations of voter fraud in the democratic primary were resolved. Johnson called in attorney Abe Fortas to seek the one outcome that mattered: get the injunction overturned, fast.

Fortas devised an unconventional, but brilliant strategy – the only one that could work in the time available. He recommended identifying and appealing to the 5th Circuit judge most likely to rule against Johnson. A loss would allow Johnson to quickly appeal to the Supreme Court, where Fortas expected they could win a final, favorable decision.

Executing on the plan, a team of lawyers flew to New Orleans to dig through previous decisions by the Fifth Circuit’s judges. After analyzing the rulings and language in case after case they found their judge, the one most likely to rule against them. Properly targeted, the court fight unfolded exactly as envisioned by Fortas. Johnson went on to win the election. And Fortas? He was appointed to the Supreme Court in 1965.[1]

This early example of legal data analysis is all the more remarkable because it was performed without technology and it is rare still today. For all of the diligence and rigor that lawyers apply to their craft, it’s a practice almost entirely devoid of data analytics. Take for example this common situation: an email goes around asking, “Does anyone have experience practicing in front of Judge X?” The responses are almost invariably sparse, personal, or anecdotal. To better answer that question with data-driven insights that escape casual observation and intuition, there are a handful of emerging companies, including my own, Ravel Law, that apply rigorous analysis to millions of judicial decisions.

By applying precise, powerful technology to legal research, attorneys can find case-winning insights – like those that Fortas discovered for Johnson – on a daily basis.

One approach we use is called machine learning, which involves training software to do something by giving it examples to learn from. For instance, machine learning can be used to identify when a judge cites an opinion and then to determine whether that citation shares similarities with how other judges have cited the same opinion. This means that what used to take days to glean, if it was even possible, can now be done in minutes.

At Ravel, teams of lawyers create training examples for our algorithms, and this combines the expertise of attorneys with the speed and repeatability of computers. Not long ago, for example, we heard from a lawyer in Michigan who used Ravel in preparation for oral argument to discover an important opinion that her team had not previously noticed. The opinion led them to adjust their argument and helped them win the case.

Data analysis and technology are opening new windows into judges’ thinking. As the saying goes: A good lawyer knows the law, a great lawyer knows the judge. Using data analysis to know your judge may be unconventional today, but it can lead to the most important outcome: success.

[1] This story is beautifully told, and in more detail, by Robert Caro in Means of Ascent: The Years of Lyndon Johnson II.

Photo by City of Boston Archives (Flickr/ Creative Common)

Related posts

Top