January 12, 2015 Revision

BIG DATA AND THE COURTS

Jim McMillan, Principal Court Technology Consultant, National Center for State
Courts

March 2015

The new world of big data has been called a management revolution. It will require a
new breed of court leaders, managers, and data scientists with the talent, tools, and
mind set to make the best use of the data coming at them in ever greater volume,
velocity, and variety.

The world creates 2.3 zettabytes (2.3 x 1021) of data every day. Ninety percent of the
data available today was created within the last two years.1 The sheer magnitude and
velocity of this vast trove of information, referred to as big data, are mind-boggling.
Because of big data, and the accompanying new algorithms and data analytics,
managers can make better decisions more quickly than ever before. Big data is
exploited today for everything, such as stocking store shelves, routing traffic, mapping
crises worldwide in real time, decoding the human genome, and finding a mate. Private
companies, public and nonprofit organizations, and individual entrepreneurs are
exploiting big data to gain new insights and to make faster and better decisions and
dramatic improvements in their operational efficiencies. Data-driven decisions are
better decisionsits as simple as that, argue MITs Andrew McAfee and Eric
Brynjolfsson, who studied the effects of big data and data analytics on productivity and
profitability. Using big data enables managers to decide on the basis of evidence
rather than intuition. For that reason, it has the potential to revolutionize management
(McAfee and Brynjolfsson, 2012, 63).
We are just beginning to build the computational and analytical toolsdata analytics
to make sense of it all, including open-source software tools, cloud computing, data
visualization, and industry-specific techniques, such as eDiscovery software that, for
example, enables law firms to analyze documents in a fraction of the time and cost than
was possible just a few years ago (Markoff, 2011).
There is considerable hype surrounding big data, and court executives, researchers,
and practitioners are right to pay attention to this trend. Because of big data, they will
know radically more about their operations and will be able to translate that knowledge
quickly into improved decision making and performance. Those who inject big data and
data analytics into their operations will acquire a calculated boldness in their data-driven

1
decision making that is likely to be a core competency of effective court leaders and
managers long into the future.
Defining Big Data
Big data refers to both the vast amount of structured and unstructured data now
available to us and the data analytics to use it.2 What makes big data different from
the mere data we have known in the past is its volume (the amount of data so huge
that it is difficult to handle using traditional databases and software), velocity (the speed
at which data are created and accessed), and variety (the range of data types and
sources), including digital data from both traditional and nontraditional sources like
images, videos, radio-frequency identification (RFID), Internet searches, call-detail
records, tweets, ubiquitous information-sensing mobile devices, GPS signals, and
wireless sensor networks. Big data, according to Gartner, Inc., the prominent
information technology (IT) research and advisory company, are the high-volume, high-
velocity, and/or high-variety information assets that require new forms of processing to
enable enhanced decision making, insight discovery and process optimization (Beyer
and Laney, 2012).3
Mobile devices alone provide enormous streams of data associated with individuals,
activities, and locations. As locations in the physical world and digital world intertwine,
courts someday, for example, may be able to ascertain how far people and their mobile
devices are from the nearest courthouse, and with this data gain insights into the
question of access to justice (Keilitz, 2012).
Consider traditional physical bookstores compared to online booksellers like Amazon
and Barnes and Noble. Amazon tracks not only books we buy in real time, but also
other books we may have looked at online; how we navigated its Web site; how much
our purchases were associated with various promotions, reviews, and page layouts; and
what books we are likely to buy based on our past purchases. Traditional brick-and-
mortar booksellers cannot access, analyze, and use this kind of information as quickly.
One of the more dramatic examples of the use of big data is Google Flu Trends, a free,
up-to-date aggregator that estimates flu activity by tracking the number of flu-related
searches against geographical and historical flu-season data in more than 25 countries.
Google data scientists found that when people have the flu, they often search the
Internet for information on relieving symptoms like fever, body aches, and cough, and
that certain search terms are good indicators of flu activity. This real-time data is a tool
for predicting and tracking influenza outbreaks that may be more effective than
traditional warnings based on doctors reports issued by the Centers for Disease
Control. Similar approaches are being used to map natural disasters like earthquakes
and floods, and other crises, in real time (Walters, 2013, 62).
Electronic discovery (or e-discovery) is the exchange of electronically stored information
(ESI) in response to a request for production of information in a lawsuit or investigation.
ESI includes e-mails, documents, presentations, databases, voice mail, audio and video
files, social media, and Web sites. The identification, extraction, collection, and

2
production of this information using digital forensic procedures and e-discovery software
is an example of big data in the legal realm that is impacting courts.4
In the public sector, by analyzing arrest and emergency-call data, law-enforcement
agencies are using big data for predictive policing, focusing resources on specific hot
spots at specific times to respond faster and prevent crimes from being committed
(Heaton, 2012). Police also are starting to use facial-recognition software and data from
license-plate video readers on patrol cars matched with booking photos and criminal
records as investigative tools in the field.
Perhaps the most dramatic examples of how far big data can take us, which many
commentators, including Christopher Surdack, the author of Data Crush: How the
Information Tidal Wave is Driving New Business Opportunities (2014), find creepy or
spooky, is a kind of mind-reading technology called anticipatory shipping,
anticipatory computing, or predictive intelligence. Online retail giant Amazon claims
that it knows its customers so well that it can start shipping even before orders are
placed (Welsh, 2014). With this anticipatory shipping, Amazon will send products that it
expects customers will want before they order, solely based on patterns of previous
orders and information it gleans from product searches, wish lists, shopping cart
contents, returns, and online shopping practices. Google, Apple, and Microsoft have
launched anticipatory computing systems that use personal data in your smartphone to
make timely suggestions, such as calling up a timetable while you are waiting at a bus
stop, offering to e-mail other attendees of a meeting that you are running late, giving
unprompted information about how long it would take to drive to a destination in current
traffic, and reducing your anxiety over missing your connecting flight by telling you that it
is also running late. Experts predict that this technology is going to be everywhere
(Standage, 2014).
The question is whether courts can exploit big data in similar ways.
Big Data Opportunities for Courts
Big data is all about joining data sets never before joined, and asking questions never
before asked and answered in real-time or near real-time. More than 96 million cases
were filed in the 50 states, the District of Columbia, and the Territory of Puerto Rico in
2012 (Court Statistics Project, 2014). Courts today produce massive amounts of data in
docket lists, judgments, calendars, Web-site-usage data, telephone records, general
case management and performance measurements systems, and electronic
documents. These data sets could be tapped using data-mining tools and then cross-
referenced against other publicly available databases to gain valuable insights for better
legal (e.g., bail and sentencing) and managerial (e.g., workload assessment and
resource allocation) decision making.
An example of a big data solution for verifying the identity of self-represented litigants
who are e-filing their cases might be something akin to the Indiana Department of
Revenues identity management system.5 In the past, an applicant for benefits would
need to visit a government office with identification, such as a drivers license or
passport. Without that personal connection and verification, identity theft and fraud

3
increased dramatically as online applications for benefits grew. One possible solution is
suggested by how selected pieces of information of a taxpayers return seeking a refund
are compared against LexisNexiss ID management service, which includes billions of
identity records in public and private databases. Could a similar solution help verify the
identities of self-represented litigants using e-filing?
Geographic information system technology (GIS) using big data enables spatial
representation of data and geography through maps. By connecting databases to maps,
GIS helps make intricate and abstract relationships visible and helps organizations and
countries to analyze and manage their systems, data, and operations more effectively.
Another use of big data in courts might come from joining structured data sets of case
filings (e.g., domestic relations) with structured and unstructured data sets of various
socioeconomic data (e.g., poverty, crime), and then using GIS to conceptualize,
visualize, and analyze possible relationships.
To reap the benefits of big data, talent matters. Over time, courts will need to employ
managers, data analysts, IT specialists, and other professionals who are comfortable
working with large quantities of information, much of it messy and unstructured data, to
discover patterns, linkages, and correlations. But correlation is not causation. Court
professionals will need to use new data analytics, as well as traditional statistics, and to
design small academic-style experiments and other inquiries to close the gap between
correlation and causation. The people who can do this will be a new breed of
professionalsdata scientists and like-minded executives and managers who believe
in the power of datawho are comfortable with big data, can speak the language of
court administration, and can help court leaders set goals, define what success looks
like, and ask the right questions to tease value from big data.6
Getting Started with Big Data
Building the capacity to use big data need not require huge investments in IT, large
blocks of time, and an army of on-site data scientists, though it does demand
technology tools and skill sets that may be new to most court analysts and IT
departments. Data analytics can be done with software tools commonly used as part of
advanced analytics, such as data mining and predictive analytics. Newer technologies
associated with big data analytics include Hadoop, MapReduce, and NoSQL databases.
Much of the needed software is open source. For example, Apaches Hadoop, the most
commonly used framework, allows for distributed processing of large data sets across
clusters of computers using simple programming models.

To get started on big data, McAffee and Brynjolfsson (2012, 66) suggest a sensible
four-step approach courts can take to explore how big data may help drive better
performance:
Step 1: Select a Testing Ground. First, a court should select a division, unit, or
specific business process of the court (e.g., calendaring, document management, or
management reporting) as a testing ground. The leaders and managers of this unit or
process should be willing to embrace big data, to explore, and to experiment. They

4
should be backed by a data scientists or like-minded professionals who are comfortable
with big data and data analytics and willing to learn more.
Step 2: Identify Opportunities for Big Data. Second, court leaders should challenge
the leaders and managers of the division, unit, or process to identify three opportunities
of big data that they could prototype within five weeks with a team of no more than five
people. The team might begin by posing an exploratory question that needs to be
answered. For example, there may be a question related to what is perceived as a low
rate of collection of fines and fees across various collection categories (e.g., criminal
fines) and collection points (e.g., a probation department). The team might recognize
that if it had a more complete big data picture of the connections between the
assessments and collection, it might be able to improve performance.
The team should first understand the data they already have, where it exists, and in
what form, including traditional and nontraditional sources within the court; within the
greater local, state, and national justice system; and outside of the justice system.
Traditional data sources include automated-case-processing information; financial
management databases; document management systems in courts, court clerk offices,
prosecutor and public defenders offices, and probation departments; fine and fee
transaction data; and databases in state criminal history repositories and federal
agencies. Less obvious sources include court-user-traffic data collected at security
checkpoints, court-user-location and activity data from smart phones, and Patchwork
Nation, a demographic and geographic breakdown of the nation into 12 different kinds
of communities using counties as building blocks.
It is inevitable that courts must prepare for the big data produced by e-discovery
systems. Courts do not have the ability to receive this data as evidence and, therefore,
must consider approaches that allow recognition and connection to these systems as
part of the trial resources and even as part of the record. One small part of this solution
has been provided in the Oasis-Open LegalXML Electronic Court Filing standard that
allows for links to data to be electronically filed into the court. Courts must also be
prepared to oversee e-discovery via court rules and processes.
The team should ask themselves the same questions and think creatively about data
collected by others outside the justice system that might provide insights into strategy
and operations, including: Is this data accessible for analysis? Does the court or court
system have the talent, tools, and technology to exploit it? If not, how can the court
acquire them quickly?
Step 3: Innovation. Third, the team should lead the court in a process of innovation
that includes discovery, experimentation, measurement, and replication. The team
should work to uncover hidden patterns, unknown correlations, and other useful
information in the data. It should systematically evaluate a variety of internal and
external data sources and conduct analysis to answer the exploratory question that the
division, unit, or business process wants answered clearly and decisively. The team
might need to ask additional questions based on what it uncovers.

5
Step 4: Sharing Results. Finally, the court should open up the resulting data sets and
analytical challenges of the three opportunities the team prototypes, not only to others in
the court but also to interested people outside of the court for comment and reactions.
A New Breed of Leaders and Managers
The technical and other challenges of big data like security and privacy issues,
particularly difficult for courts, are very real, but the managerial challenges may be
greater. As suggested by Professors Donald A. Marchand and Joe Peppard (2013) in a
recent Harvard Business Review article, Why IT Fumbles Analytics, deploying
analytical tools is relatively easy. How to discover and to use the information these tools
help generate to make better decisions and to gain insights is much harder. Big data
and data-analytics projects should focus less on the deployment of technology, they
counsel, and more on the exploration of the use of information to solve business
problems.7
First and foremost, court leaders, managers, and other court professionals will need not
only new skills but also a new mind set and management style to see information not as
objects residing in databases and dashboards but, rather, as outcomes of people
constructing meaning out of data. A new and different breed of leaders and executives,
with new skills and management styles, are likely to gain prominence in the courts in the
future. They will embrace the imperative of performance measurement and feel
comfortable with the challenges and opportunities of the vast new streams of
information available to them.
McAffee and Brynjolfsson (2012) suggest that executives interested in big data develop
the following discipline. First, they should develop a culture in which their first question
is not What do we think? but rather What do we know? and What dont we know?
When facing a decision, they should get into the habit of asking What do the data say?
and then follow up with questions such as Where did the data come from? and How
comfortable are we with the results?
Second, they should develop a mindset that allows themselves to be overruled by the
data when the data do not agree with their hunches or intuition. The evidence is clear,
say McAffee and Brynjolfsson. Data-driven decisions tend to be better decisions.
Leaders will either embrace this fact or be replaced by others who do.

Reports are part of the National Center for State Courts' "Report on Trends in State
Courts" and "Future Trends in State Courts" series. Opinions herein are those of the
authors, not necessarily of the National Center for State Courts.

6
References
Beyer, M. A., and D. Laney (2012). The Importance of Big Data: A Definition. Gartner,
Inc., Stamford, Connecticut.
Heaton, B. (2012). Behavioral Data and the Future of Predictive Policing. Government
Technology, November 2.
Keilitz, I. (2012). Big Data, Data Analytics, and the Access to Justice Card. Blog,
Made2Measure, November 4.

Marchand, D. A., and J. Peppard (2013). Why IT Fumbles Analytics. Harvard

Markoff, J. (2011). Armies of Expensive Lawyers, Replaced by Cheaper Software.

McAfee, A., and E. Brynjolfsson (2012). Big Data: The Management Revolution.
Harvard Business Review, October.

Court Statistics Project (2014). Examining the Work of State Courts: An Overview of
2012 State Trial Court Caseloads. Williamsburg, VA: National Center for State Courts.

Surdak, C. (2014). Data Crush: How the Information Tidal Wave Is Driving New
Business Opportunities. New York: AMACOM.

Standage, T. (2014). When Smart Becomes Spooky. Economist, November 18.

Walters, P. (2013). Crisis Mapper. National Geographic 223(2).

Welsh, W. M. (2014). Amazon Says It Can Ship Items Before Customer Order. USA
Today, January 16.

Endnotes

1
Big Data, Wikipedia.
2
Big Data, Webopedia.
3
This 3Vs definition can be traced to a February 2001 research note, 3-D Data Management:
Controlling Data Volume, Velocity and Variety, by D. Laney, research vice president for Gartner
Research; see also, P. Miller, (2012). The Next Big Thing: Wee Data, Blog, Cloud of Data, September
25, 2012.
4
J. Bertolucci, When Big Data Meets Legal Discovery, Information Week (March 19, 2013).
5
J. Moore, Indiana Combats Tax Refund Fraud with ID Verification, Government Computer News
(January 27, 2014).
6
See T. H. Davenport and D. J. Patil, Data Scientist: The Sexiest Job of the 21st Century, Harvard
Business Review (October 2012): 70-76, noting that data scientists are professionals with training and
curiosity who know how to fish out answers to important business questions from big data.

7
7
In a January 30, 2013, Harvard Business Review webinar, Big Data: Can You Seize the Opportunity,
based on their article, Marchand and Peppard noted: Business intelligence does NOT reside in the data
warehouse or with big data. Rather, BI emerges in the minds of employees when they identify and access
data, combine data with their own or others knowledge of a business situation, and produce novel
insights and/or resolutions to the issue at hand.