Alex Jakubow discusses Duke Law's enhanced empirical research support capabilities

August 26, 2019Duke Law News

Alex Jakubow recently joined Duke Law as associate director of empirical research and data support services, a new position based in the J. Michael Goodson Law Library where he oversees a team supporting empirical research and scholarship. Prior to moving to Duke he was empirical research librarian at the University of Virginia School of Law, where he co-directed its Legal Data Lab, a database resource that includes the Corporate Prosecution Registry now maintained in collaboration with Duke Law.

Jakubow has worked as an applied statistician and social scientist and also taught classes in political science and research methodology at several universities including UVA. He earned his PhD in political science from Rutgers University and holds a dual BA in history and political science from the University of Richmond.

“Alex comes to Duke Law with deep expertise and many years of experience as an applied statistician and social scientist,” says Femi Cadmus, the Archibald C. and Frances Fulk Rufty Research Professor of Law, Associate Dean of Information Services and Technology, and director of the Goodson Law Library. “I am excited to see him lead the library’s work in supporting our faculty’s expanding data research and projects.”

Jakubow spoke with Duke Law Communications about the new Empirical Research and Data Support Services team and why lawyers should become familiar with empirical methodologies and analysis.

Q: What are your goals for this new department?

Our department's primary duty is to support empirical legal scholarship at the Law School. We have strong methodological foundations in a variety of statistical procedures and inferential techniques commonly used in the social sciences. Coupled with our competencies in general computing, web-oriented, and statistical programming languages, we also create the datasets our clients need to test their hypotheses and more generally advance their own research agendas in a particular area.

Going forward, my hope is that we can marshal additional resources to broaden the network of individuals and groups we regularly serve, as well as to support emerging scholarship as it applies to the extraction, ingestion, and analysis of legal documents using natural language processing, machine learning, and other advanced computational techniques.

Second, we aim to multiply the impact Duke Law scholarship has on the broader academic, public, and professional communities. Similar to what we did for the Corporate Prosecution Registry at University of Virginia (now jointly maintained with Duke Law), we can help faculty members develop companion sites and hosting domains for their scholarship. These digital, public spaces allow our researchers to go beyond the confines of traditional academic publishing and experiment with different ways of sharing their knowledge with the world. At the same time, we hope to facilitate greater awareness of digital preservation. Hosting the datasets, code, and other peripherals used as part of the publication process in one or more institutional repositories is not only good scientific practice, but it pays huge dividends to researchers — both in terms of workflow efficiency and research impact.

This relates to a third goal involving scientific rigor in legal empirical scholarship. Other social science disciplines have done more to ensure that published research reflects sound procedural practice and is otherwise replicable. Barring a few exceptions, legal scholarship has lagged in this area. This presents a real opportunity for us. Promoting the use of repositories — such as Duke’s own Research Data Repository — would certainly help. Another part of this strategy could involve working with our student journals to develop sound criteria for evaluating and publishing empirical manuscripts. I would love to see Duke Law become an exemplar of transparent, replicable research among our peers.

Finally, I hope this department — and the law library more generally — will be able to play a more formative role in empowering our students to be critical consumers of empirically-oriented work. The classic line that those "who don't like/know math go to law school" no longer fits with the emerging reality of modern legal practice. For better or for worse, our students are going to be increasingly exposed to data-oriented arguments inside the courtroom (and beyond). The ability to differentiate between valid and invalid claims can be vital for success, and this type of critical thinking would be a great way that Duke grads could set themselves apart from the rest of the field.

Q: Femi Cadmus, immediate past president of the American Association of Law Libraries, has said that the ability to analyze and visualize data is an increasingly important skillset in law librarians. Why is that?

First, the general march of quantification through academic, commercial, professional, and public domains has catalyzed exponential growth in the overall availability of data about the world around us. Coupled with staggering advances in computational power and methodological techniques to study all of this new data, people can get overwhelmed. Data-savvy librarians can help patrons identify appropriate data sources for their research agendas, keep abreast of new data sources for potential use, and transform data from a variety of formats with which folks outside of librarianship may be unfamiliar — such as MARC (machine-readable cataloging) records, JSON file formatting, HTML — into something they can use more intuitively, such as tabular datasets or spreadsheet documents.

Second, adding data skills to the law librarian toolkit allows us to take advantage of our institutional position within law schools in a way that supports the documentation and discovery of empirical projects. Libraries are repositories of knowledge, and the raw inputs of empirical research — datasets, codebooks, program files — are also pieces of knowledge deserving of preservation. Libraries are a natural home for these materials, and we already have the infrastructure to catalog them.

Finally, law librarians are deeply embedded in the faculty research workflow. Because of this, we have a good sense of how best to situate the researcher's intent or problem within the broader substantive context and use the resources available to us to figure out what data sources, analytical techniques, and/or data deliverables might be most useful for a given project.

Q: You collaborated, when he was at UVA, with Brandon Garrett, the L. Neil Williams Professor of Law, on the corporate prosecution registry, as well as a paper on the decline of the death penalty in the U.S., and a paper on risk assessment. What else attracted you to Duke Law?

I am excited to be working with him again, and for all of the wonderful opportunities here at Duke. I am struck by the refreshing interdisciplinary approach the community at Duke tends to take when tackling important social problems and conduct research. If you have a useful skill set and if you have something to contribute — irrespective of your job title or home department — you can be part of the project. This is a welcomed deviation from the longstanding norm of siloed research common in many other institutions of higher education.

There are a lot of great data resources here, such as the Center for Data and Visualization Services, the Social Science Research Institute, Research Computing, the Rhodes Information Initiative, and Data Analytics and Practice — to name just a few. Part of what I’m hoping to do here is establish and build good relationships with these other service units around campus. In a field as dynamic and expansive as this one, it’ll be rare that one person will have the relevant depth of knowledge or skill sets appropriate for all data needs, so it’s all the more important to foster collaborative partnerships and be able to build on the comparative advantages and relative strengths that different stakeholders have.