The First Round of the On Think Tanks Data Visualisation Competition was won by the Mapping Arms Data visualisation produced by the Igarapé Institute of Brazil in collaboration with Google Ideas and PRIO. Submissions for Round 2 of the competition are open until 2 October 2013, and I wanted to find out more about how this technically advanced, visually stunning and information packed data visualisation came into being with the hopes of inspiring other think tanks to consider putting together their own (maybe not so advanced!) visualisations. As such, I sat down with Robert Muggah, one of the visualisation's creators, for an interview.
Jeff Knezovich: Let's start at the beginning, where did the idea for Mapping Arms Data come from?
Robert Muggah: The MAD tool was originally conceived for a major conference organized by Google Ideas in 2012. The Igarapé Institute was approached to explore the connections between authorised and illicit small arms flows. The tool itself was created on the basis of extensive interaction between designers at Google Ideas and subject matter experts at Igarapé and the Peace Research Institute in Oslo (PRIO). In addition to launching the tool, the Igarapé Institute and PRIO also used the MAD tool to generate awareness during UN negotiations on small arms control and an arms trade treaty throughout 2012 and 2013.
JK: Was the visualisation always integral to the idea, or did it emerge after you had a lot of data?
RM: The original visualisation was feasible because Igarapé Institute and PRIO had already assembled considerable amounts of data. Indeed, PRIO´s Norwegian Initiative on Small Arms Transfers (NISAT) has been collecting, cleaning and refining data on small arms, light weapons, and ammunition for more than a decade. NISAT´s primary data source is the UN commodities database (or COMTRADE), but the research project also features information from over 40,000 articles on exports and imports. Both Igarapé and NISAT had long been interested in identifying intuitive and compelling ways of exhibiting complex data, and the partnership with Google Ideas allowed this idea to come to fruition. The process in developing MAD was comparatively straight-forward since the data was available and relatively “clean” from the get-go.
JK: The MAD visualisation is very technically advanced, what sort of technological challenges did you face in producing it?
RM: While technically advanced, the development of MAD occurred comparatively quickly. The basic platform for the original toolwas WebGL software. The construction of the visualization itself required rapid iteration between the Igarapé team and a programmer affiliated with Google Ideas. In order to render the data more accessible to non-expert audiences, we collapsed the many types of arms and ammunition into three simple generic categories (in the latest version of MAD, we added in a fourth category of armaments – "unspecified"). Additional apps were also designed to allow for geo-spatial contours to be added. Finally, a separate team of experts was recruited to add bar graphics, filtering options, a timeline scrubber and historical graph.
JK: Sounds like it was something of a team effort! Did you go out seeking technical advice and expertise, or how did the partnerships come into being?
RM: The MAD tool was developed on the basis of a partnership between Google Ideas, the Igarapé Institute and NISAT. Google Ideas and Igarapé Institute had collaborated previously on a modest initiative exploring the relationships between popular film and illicit networks. Igarapé Institute and NISAT personnel had also collaborated extensively for more than a decade on a range of issues connected to arms transfers and armed violence. All of the technical skills thus resided “in-house” – with Google Ideas providing engineering and product design expertise and Igarapé and NISAT offering technical and data-related support. It is worth emphasizing that Igarapé and NISAT provided content while Google Ideas facilitated its visualization.
JK: The visualisation was published about a year ago, but how long did it take to actually develop?
RM: The original small arms and ammunition visualisation took about six months to develop starting in early 2012. The project was developed on the basis of a basic agreement between the parties with the intention of launching it in 2012. It was later featured in Beijing, New York, and Rio de Janiero at various launches in 2012 and 2013. It also received considerable media attention. An updated version integrated new data (for 2011, the latest year for which information is available), refined some earlier bugs in the program and dataset, and introduced some new “cleaner” features. The upgrading took another six months and the new tool was re-launched in May 2013 in Villingen-Schwennigen, Germany.
JK: We've always suggested here at On Think Tanks that communication activities -- like creating data visualisations -- are only a means to an end, and not an end in themselves. You've already hinted at it above, but how have you put the visualisation to use?
RM: The original small arms and ammunition visualisation was launched with Google Ideas in Los Angeles and received considerable attention. It was featured on the Google Ideas website, as well as by the Igarapé Institute and PRIO. Within months of its launch the tool was viewed by over a million people from government, business, the tech world, the multilateral and not-for-profit world, and academia. It was picked-up in the Atlantic, CNN, Forbes, Huffington Post, the AtlanticWire, Mashable, and literally hundreds of media outlets in 12 languages. In addition to winning a few awards, the video of the launch was also put up on Youtube and collected almost 20,000 views. We also distributed the tool actively through the arms control and disarmament communities in and outside of the United Nations to shape debate during negotiations on small arms as well as a larger arms trade treaty (which was agreed in May 2013). The updated MAD tool was launched in 2013 at a major conference with op-eds featured in the Atlantic and stories in hundreds of conventional and social media outlets, including the BBC. By September 2013, it had been viewed over 3 million times in over 150 countries.
JK: What do you think the benefits of visualising the data in this way have been for your research?
RM: Ultimately, the purpose of MAD is to make the arms trade more accessible and legible to a wider audience. By presenting a large dataset in visually arresting and user-friendly manner, it has inspired “mainstream” debate, but also people associated with technology and design industries, police and justice, relief and development, and beyond. Intriguingly, it was also shown in China at the Beijing Design Week in September and October, a country not typically known for its openness on arms-related issues. The collaboration between Igarapé, PRIO, and Google Ideas is a reminder that technology is not just an add-on, but increasingly a central part of content development and messaging. Researchers and practitioners will need to engage and adopt many of these visualisation and analysis tools – including on issues such as the arms trade – if they are going to improve their work and trigger policy change.
JK: On the flip side, have there been any particular drawbacks for you having this represented as a visualisation? If so, what have they been, and how have you worked to overcome them?
RM: We have yet to see any major drawbacks in developing MAD. Of course, in some countries the debate on small arms and ammunition is volatile. It is often informed and fueled by ideology over evidence. We have received some critical feedback from the more extreme periphery of the pro-gun movement (owing in large part to the major exposure provided to the MAD tool among gun owners). However, what has been important is that the vast majority of those on either side of the gun debate have welcomed the tool, often reaching out to ask more questions and retrieve more information. On balance, most users recognise that MAD is a transparency device, and intended to shape more educated debate on the wider dynamics of small arms and ammunition trafficking. More explanations of the MAD tool and its implications for research and policy are available here or here.
JK: Many thanks for your time Robert, and congratulations again on winning the first round. We'll be interested to see how the visualisation fares in the final!