Write a response

Ethical questions in data journalism and the power of online discussion

Developing ethical standards for working with data


Data journalism projects from different topics and locations are united by the ethical challenges they present.

As scholars and practitioners of data journalism have pointed out, main issues include flawed data, misrepresentation from a lack of context, and privacy concerns. Contributors have discussed the ethics of data journalism on this site in posts about topics such as transparency about editorial processes in computational journalism and best practices for doing data journalism ethically.

Our research project looked at similar ethical challenges by examining journalists' discussion of the controversial handling of publicly accessible gun permit data in two communities in the United States. These cases are from the early 2000s, but the issues they raise persist and signal opportunities to learn from the online discussion of ethical issues, and to ask a wide range of ethical questions about data journalism.

Rawpixel pexels

The cases

Less than two weeks after the 2012 shooting and deaths of 20 children and six staff members at Sandy Hook Elementary School in Newtown, Connecticut, a journalist at The Journal News in White Plains, New York, wrote a story about the possible expansion of publicly accessible gun permit data. The article was accompanied by three online maps with the locations of gun permit holders. These clickable maps of a two-county area in the New York suburbs also included the names and addresses of the gun permit holders, which allowed readers to identify them. The inclusion of this personal information prompted a public outcry both locally and nationally, mainly relating to privacy and safety concerns, and the maps were subsequently taken down.

POD1bigger

Media outcry over The Journal News' map.

Although the 2012 case prompted the greatest attention, another New York reporter’s Freedom of Information request for a database of gun permit holders in three counties sparked an earlier public outcry in 2008.

In response, the newspaper’s editor published an editorial: “We here at The Post-Star find ourselves in the unusual position of responding to the concerns of our readers about something that has not even been published in our newspaper or Web site”. The editor said the request “drew great concern from members of gun clubs and people with gun permits in general, a concern we totally understand”.

POD2

NICAR-L offers journalists the opportunity to discuss all things computer-assisted reporting.

Both of these cases prompted discussion among journalists, including participants in NICAR-L, the listserv of the National Institute for Computer-Assisted Reporting, whose subscribers include data journalists from major news organisations in the United States and around the world. Our study examined the content of three discussion threads -- a total of 119 posts -- that focused mainly on ethical issues.

Rawpixel pexels

Key ethical issues

Several broad ethical issues, and specific themes related to those issues, appeared in the discussion.

1. Freedom versus responsibility and journalistic purpose

One of the most frequently mentioned ethical themes was whether gun permit data should be posted online just because they are public data and can be posted. “Yes, it is public information but so are a lot of things that are not ethical or responsible to publish,” wrote one lister. Many listers said the data should serve a broader public purpose to be posted.

2. Privacy and verification

The two concepts are linked because publishing inaccurate information can lead to invasions of privacy, and journalists have a duty to verify it before it is published. The main theme was that journalists should consider the risks to a person’s private life when deciding whether to post data online. One lister wrote, “And if we’re going to make something public, or publicize that it’s public, then we need to weigh the value of publishing against the risks posed to the people included in the story/database/graphic.”

3. Consequences

A substantial amount of discussion related to possible consequences of publishing the maps and database, mostly negative but a little positive. One of the most prominent themes was that gun permit holders and their families may be at risk from thieves because their names and addresses were posted online, and their guns are valuable. A few posters thought knowing someone had a gun could deter thieves, who might fear it would be used on them. Another concern was that posting such information could cause government officials to take away access to this data. In fact, the state legislature quickly passed a law restricting journalistic access to the data.

4. Alternatives

The alternatives frame focused on possible approaches that might avoid or minimise ethical problems. These approaches included trying to notify gun permit owners that their names and addresses will appear online -- difficult as that may be -- and using aggregate data instead of individual gun permit data to avoid showing individual names and addresses, and instead indicating trends.

Rawpixel pexels

Questions to ask when making ethical decisions about data

The ethical considerations raised in this online discussion helped us develop the following set of questions that journalists should ask themselves before they publish:

  1. Does the information serve a journalistic and public purpose? To what extent? The data must at least serve both of these purposes to be posted online.
  2. Who could be harmed by the information? To what extent? Are there risks to a person’s private life from any elements of the data? What is the potential impact if data is erroneous or out of date?
  3. Are there alternatives that would maximise the public purpose, such as combining it with information from other databases? Are there alternatives that would minimise harm, such as aggregating personal data instead of using individual names and addresses?
  4. Can the data be verified? Have reasonable steps been taken to verify the accuracy of the data? Can people in the database be notified before publication? What can be done to enable correction of data errors identified after publication, including updates because people have moved?

We believe that the final decision should take all of these considerations into account, and prioritise journalistic purpose and harm minimisation.

Rawpixel pexels

The value of online communities in discussing ethics

Beyond the questions that emerged, our study underlines the value of online discussion in developing ethical data journalism practices.

In the perspective of philosopher Alasdair MacIntyre, who wrote about the role of social practices such as professions, the members of NICAR -- mostly current or former practitioners of data journalism or related work -- are distinctively qualified to offer nuanced insight into the ethics of data journalism because they are bearers of the standards of excellence of the practice. As media ethicist Sandra Borden has written, the practice functions as a moral community, and one of the ways this happens is through “discourse that sustains, repairs, and extends standards of excellent journalism”.

The discussion on the listserv sustained these standards of excellence by highlighting longstanding ethical considerations that continue to be relevant as journalists increasingly work with data. It repaired standards in the sense that participants critically evaluated the ethics of the decisions made. The discussion did not, by itself, extend ethical standards in data journalism. However, these threads represent an important place for the development of standards and alternative proposals for better practice.

Rawpixel pexels

The content of this post has been adapted from David Craig, Stan Ketterer, and Mohammad Yousuf’s research: To Post or Not to Post: Online Discussion of Gun Permit Mapping and the Development of Ethical Standards in Data Journalism in Volume 94, Issue 1, of the Journalism & Mass Communication Quarterly (pages 168-188). Copyright © 2017 by AEJMC. Used by permission of SAGE Publications, Inc.

subscribe figure