I’m a big fan of the Stockholm International Peace Research Institute’s Arms Transfers Database. SIPRI makes it extremely easy to spit out highly legible records of military equipment transfers between nations and non-state groups. I’m not an expert in such matters, but the database is approachable enough that I can readily find, say, US sales of cluster bombs to Saudi Arabia. Legibility and approachability, however, do not mean the data are machine-friendly. SIPRI exports data in a rich text format, filtered by year, exporter, importer, and class of equipment. The exported files are superficially tabular in arrangement, but do not paste well into a spreadsheet without some cleanup. I wanted to visualize trends in US arms exports from 9/11 to the end of the Obama administration, particularly focusing on client states that are now bombing civilian targets in Yemen. To that end, I would need to reformat some data from SIPRI.

I downloaded trade registers for all weapon types from 2001 to 2016, with the recipients unfiltered and the USA as supplier. The resulting .rtf copied into Excel better than I expected, but I needed to realign some columns offset by tabs in the original text. It was also important to fill out the “recipient” column so the appropriate recipient name recurred for every transaction–not just as a section header.

After reaching the limits of what could best be accomplished in Excel, I performed some basic clustering in OpenRefine to make sure my CSV would be as clean and uniform as possible. This mostly entailed dealing with some inconsistencies in the “weapon description” column. I would use this later to group weapons into broader categories.

Then, as per the parameters of the assignment, I loaded my data into Tableau Public and began to explore. While I suspected the arms transfers dataset would reveal interesting patterns, I wasn’t yet sure how to find them. I definitely wanted to visualize change over time, but for starters a simple bar chart without a chronological element seemed like a good idea. Sorting recipients by total number of items delivered immediately showed a preponderance of arms flowing into the Middle East, particularly Saudi Arabia and Israel. Others, like CNN, have used SIPRI’s data to create much the same chart:

NB: My chart and others’ stabs at the same share a major issue. They rely on SIPRI’s count of number of items in each transaction, as opposed to the value of the items transferred. This method of comparison makes sense when comparing a specific class of item (guided bombs for example, see below), but falls apart when a single bomb and a multirole fighter jet are given equal weight.

Exploring change throughout the “War on Terror” would mean bringing time into the mix. I opted again for the simple and familiar option: line charts. Mother Jones demonstrated an alarming uptick in US arms sales in 2011 thus:

Their chart gets the point across pretty well. It shows the US blasting out of the cluster of other nations in 2011. True to the publication’s mainline American liberal stance, it implies that a certain baseline level of arms dealing as normal, and the US stepped out of line. I think it’s a little cluttered, but it tells a story. Mother Jones also used an area chart to demonstrate a Wealth-of-Nations-ass narrative of the US taking other nations’ market share. How dare we!

The jagged ups and downs of these charts don’t do it for me. Simply charting sales year to year doesn’t communicate the buildup of arms and equipment over time. Bombs sold to a Gulf autocrat in 2008 may well remain in his arsenal until needed years later. Combat vehicles will almost certainly see years of use. History has demonstrated that arms flowing into a conflict zone will stick around for decades. The Ardennes Forest still contains unexploded ordnance from both world wars. Nazi assault rifles have seen use in Syria alongside scads of mid-century Kalashnikovs. I opted to chart the running sum of certain arms sales in order to give a sense of this flooding effect.

As I mentioned above, the variety of equipment SIPRI tracks necessitated some thought and grouping. I charted “munitions” as a broad category (excluding vehicles, radar, anything not made to explode or launch exploding things) and guided bombs as a specific one. The guided bomb chart is designed to illustrate the accelerating sale of bombs to the Saudi-led coalition that’s been pulverizing Yemen. The coalition nations are catching up to Israel, and the combination of the coalition and Israel outstrip the rest of the world in deliveries of US guided bombs. A significant share of human misery in the Middle East over the coming years and decades will be of our making.

The more general munitions chart shows Saudi Arabia leaving all other players in the dust as a recipient of American weapons–even Israel, darling of neocons. I hope this visualization conveys that the real beneficiaries of US foreign policy are Likudnik fascists and Wahhabi plutocrats.

I kept my dashboard of three visualizations very simple, taking care to reuse colors where applicable. I selected colors on the line charts based on the dominant colors of some nations’ flags: green for Saudi Arabia, blue for Israel, brown as a blend of Egypt’s red, black and gold. Annotation and interactivity are minimal.

There’s plenty of room for expansion and improvement here. I’d like to explore other chart types available in Tableau, as well as its capacities for interaction and embellishment. At the risk of getting even more didactic, a little illustrative “chart junk” will be fun to play with in the future.