Representation Matters

Diagrams representing the relative Mortality from ZYMOTIC DISEASES, from WOUNDS, and from ALL OTHER CAUSES in the HOSPITALS of the ARMY in the EAST

Nightingale returns in 1859 and “corrects” the batwing diagrams with the diagrams above, now known as polar area charts or Nightingale roses. This is more “correct” in that the one-dimensional variable of mortality rate is reproduced as the single variable of area, meaning that the wedges are appropriately proportionate to each other. However, unlike with the stacked bar chart, here area is measured from the center, meaning part of many of the purple wedges is covered up by orange and green. So while, still, zymotic disease remains a runaway killer, now it’s, oddly, underrepresented. Furthermore, expecting people to accurately compare areas as opposed to lengths is a tall order. We can tell which line is longer than which, and typically we can tell about how much longer (two times, three times). With area, it’s a bit trickier.

Which circle is two times larger than the leftmost? And how much larger is the other circle? It’s hard to tell, since we can’t visually stack the circles, unlike the way the bars can become mental Cuisenaire rods. In both visualizations—batwings and roses—Nightingale gets an effect from the fuzzy visual area calculations. In the first set, the result is massive distortion in favor of her argument. In the second, it’s a slight distortion against.

But it is a fiction to assume that one rendering is more “correct” than the other; rather, the question is what distortions are within the realm of the reasonable in terms of making the narrative and persuasive point one wants to. As Mark Monmonier explains in opening How to Lie with Maps, maps are only readable and coherent because they have distortions. As Lewis Carroll, Jorge Luis Borges, and Umberto Eco (among others?) remind us, if maps were “true,” they would be unusable.

Volume 1

The first volume of Torn Apart / Separados challenged us with several questions in terms of visualization, some of which were then discussed in several of the first reflections. There was a temptation, for example, to size the various dots in proportion with certain data associated with them, such as average daily population of detention facilities used by ICE.

But that would have distracted somewhat from the other, competing desires. If the goal is to show that ICE is everywhere, then scaling the markers based on average daily population will make some facilities jump out and others disappear into view. The story becomes different and more focused. In “Clinks” and “Charts,” by making all the in-use facilities look the same on first blush, the banal repeatability of ICE as it infects our national body seems more thorough.

On the other hand, in “Banned,” we rely on cartographic distortion to overstate a case. This visualization draws a map of contiguous United States whose combined population is close to but less than the approximate number people banned from entry into the United States at the moment, blocked by the upheld “Muslim Ban.” Yet as sparsely populated mountain or plains state after state is added to the growing black shape, it soon seems like nearly all of America would be banned. We trick the eye, then, into thinking that American population is evenly distributed across the United States. But it isn’t.

This is, unfortunately, a common design choice in choropleth maps, and it resembles the “error” in Nightingale’s batwing, as well. I, at least, would not make such a map under normal circumstances.

But these are not normal circumstances. In this way, “Banned” is a response to the famous map of the 2016 Presidential Election hanging in the West Wing, as tweeted by Trey Yingst:

So just as fewer people voted for the red candidate than the blue, “Banned” exaggerates the size of the US that would be banned from entry.

The red candidate understands that this map is making an argument, that the totality of red space gives the visual illusion of an America standing behind him as President, as the true voice of the majority, and of an overwhelming majority at that.

Volume 2

The second volume, however, also presented challenges in terms of representation. “Lines,” for example, shows how pervasive the American removal engine is. Yet the number of deportations at a specific port of entry can be single digits or it can be over 1,000. Furthermore, simply drawing lines devalues the fact that each individual removal is its own story that’s the most important in the world to the deportee and their family and loved ones. But somehow the difference between one and 1,000 also needs to be indicated. The solution to me was to use a logarithmic scale. Now, 1,000 is a line that’s four times longer than the line for a single removal, and every soul gets a wedge cutting across the face of the United States as it continues to expel.

In this way, “Lines” has it both ways. It shows the extent of the problem of removals across the world as a whole (after all, the US has no problem removing souls from points of entry like Ireland), but also makes clear that even one removal is a scar on the national image.

The Banality of ICE

Both Freezer visualizations work from literally the same data set, indicating that the same hierarchically grouped data can represent radically different modes of being and interacting with the government contracting machine.⁴ In one visualization, a messy ball of yarn shows how interconnected contractors are with the various governmental needs in the awards they win. A sprawling monster like CoreCivic provides “security guards and patrol services” (which we categorize under “The Threat of Violence”), “correctional institutions” (“Walls”), and “facilities support services” (“Surveillance”). Firms like CoreCivic, or even more abstract management and consulting firms like Booz Hamilton, have their fingers in various ICE contracting pies.

Nevertheless, there are not actually that many contractors that are that promiscuously pie-fingering. A company like Spectrum Security Services simply provides “security guards and patrol services” over and over until they have made $180m in ICE awards for 2018. In “ICE Tray,” that focus is shown in having Spectrum Security services with just one square (but a large one) inside the “The Threat of Violence” category, where all “security guards and patrol services” awards land. CoreCivic, on the other hand, has several boxes throughout the tray in several differently colored areas, to show its relatively broad reach.

In the other visualization, however, our nine invented categories of government awards are more or less laid atop each other, giving the impression of a Gordian knot of governmental impenetrability. Neither visualization is right; they just aim to highlight different aspects of the same data to make a similar, underlying argument.

In short, ICE funding can be stultifyingly banal. The outrage over the zero tolerance policy in part relies on characterizing ICE as a bunch of special forces cosplayers, kicking people of color with steel-tipped boots and throwing their children in cages. But it’s not just that, else ICE funding would go exclusively to boots and cages. It’s also the massive apparatus that keeps Hillary-voting, lanyard-wearing, Pod Save America–listening Northern Virginia technocrats fed, technocrats largely represented by Democrats in Congress.

The oversized role of these technocrats is hinted at in every visualization for volume 2 save “Lines.” In “Districts,” we see that ICE funding is not evenly distributed around the US, with 16 Congressional districts taking home almost 90% of the ICE budget since 2014. Of these 16, several are DC-adjacent, where companies like Phacil and WidePoint Integrated Solutions pull in millions of dollars from ICE despite (and because of) being, basically, IT consultants. Unsurprisingly this explosion of money to the IT sector happens to be a trademark of Representative Gerry Connolly’s work in Congress. Small wonder, then, that his district is also the most remunerated by ICE. Since 2014, $1.3B has been showered over the Clinton-by-39-points district, much of it to IT companies.

This visualization, showing the distribution of the over $9B in ICE spending since 2014, recontextualizes how those 16 districts chew up the budget. In 17th place, marked in orange above, is Washington, DC., which while another ICE fat cat, also marks the bend towards quickly shrinking turns at the authoritarian trough. And this visualization skips the quarter of all Congressional districts that haven’t seen a dime from ICE. Perhaps they need to get in the consulting, computing, constructing, and coercing businesses.

The bipartisan nature of the grotesque gluttony also expresses itself in these top districts, where the pitiful Democrats, shut out of every avenue of access in Washington, or so they tell us, still manage to bring home the fat in thicker, richer slices than their GOP counterparts.

Finally, I should in passing, tying “Districts” to other current events, note that Duncan Hunter may be woeful with his family budgeting, but his district has done well under the ICE regime, as Spectrum Security Services has brought in $860M to parts of San Diego and Riverside Counties. California’s 50th is the fourth most remunerated district.

The concentrated ICE spending also lets minority- and women-owned companies shine in the glow of government awards, as we see in “Gain.” Much of the $890M going to Alaska ends up in the coffers of businesses that are registered with the government as “Alaska Native Corporations” (ANC). One such company, Barling Bay, which is a subsidiary of the Old Harbor Native Coporation, describes the benefits of doing business with an ANC by pointing out that “Alaska Native Villages suffer from some of the worst poverty in this country,” and, hence, it’s important to support ANCs who enjoy special rights in the Federal procurement process.

Barling Bay seems to provide ICE only with IT-related services, but the reach of the biggest ANC bringing money to Alaska is much greater. The NANA Regional Corporation represents over 14,000 Iñupiat shareholders, who largely participate in subsistence activities and full-time or part-time employment above the Arctic Circle. Nevertheless, NANA’s subsidiary, Akima, has grossed those subsistence shareholders over $200M in ICE awards since 2014, mostly in relation to running the Krome detention center in Miami and the Buffalo facility in Batavia, NY. Akima’s ability to win awards for various detention-related services contribute to making Alaska’s At-Large District the third most flush with ICE cash.

Back in the Beltway, Phacil, based in Virginia’s 8th District (5th best remunerated), has the dubious distinction of being the most-remunerated “Black American–owned” (the government’s designation) business regarding ICE, which has paid it $310M since 2014 for various IT-related services. The Phacil website recalls the company’s start as “a small, minority owned business” that has, obviously, “always embraced diversity and inclusion,” but the idea that the money ICE has paid them has gone towards encouraging racial equality in the United States is an offensive joke. These minority-owned companies receiving over $1B from ICE since 2014 are, it seems, largely in the various banal businesses of logistics and IT, enriching workers already in these largely white-collar sectors. Though contractors like Akima are actively in the business of jailing their “fellow” people of color, most simply contribute to the more abstract carceral regimes of the state, facilitating the ways in which a laser printer helps keep another child separated from their parents.

Speaking of laser printers, finally, in “Rain,” we see how ICE’s awarding has grown since 2014. But the tiny dots in the visualization hide a simple detail visible only if one happens to mouse over them in just the right way that causes nearly the whole visualization to suddenly become darker. Of the 5,500 awards represented in “Rain” (and in “Gain” and “Districts,” for that matter), over 1,000 of them went to one company, North Carolina’s Net Direct Systems. The $28M Net Direct has brought to North Carolina’s 2nd district is comparable chump change on the ICE scale, but the idea that someone in the ICE office, one out of every six times, has awarded a contract to Net Direct boggles the mind and fully delineates how normal the ICE funding apparatus is.

If there is an entity to which nearly 20% of my purchases (in terms of number, not in terms money spent) goes, it would probably be a grocery store. Those trips feature the near daily purchases that provide the fuel that keeps me running. Similarly, the laptops, tablets, and desktops ICE constantly buys from Net Direct are also, then, the fuel that keeps this regime of terror running.

---

Moacir P. de Sá Pereira (@muziejus) is a scholar of literature and space.