The long, sad rabbit hole of politics, healthcare, and the Supreme Court: How I tried to draw a map and failed

Numbers are so, weird. They lack emotion, judgement, subjectivity. And yet they are reflections of ourselves. And we are so willing to manipulate them, fight over them, use them to control one another. They are the good and the bad in us. I know you know this.

But this week, I was reminded of this yet again, when I tried to count states and draw a map.

It was a really tough week. On Wednesday, the Supreme Court heard oral arguments on the King v. Burwell case. That’s a case that will decide whether the feds get to continue giving the subsidies that help people pay for healthcare in a whole bunch of states (Did you get the “a whole bunch” bit? Hold on to that). You can read about the case on Vox in a really good explainer here. Even Michael Cannon likes it (and he doesn’t like the ACA much), except for these parts.

Numbers really matter in this case, and I want to talk about them, but perhaps not in the way that you think. You might be thinking that I’m about to make the case that the people in the states at risk won’t be able to pay for health insurance if they lose subsidies, and that I’m about to show you those numbers, but I’m not. That’s my day job.

How many states would lose subsidies if the Supreme Court decides in favor of the plaintiffs in King v. Burwell?

I want to talk to you about numbers in a different way. Here they are:

  • 37
  • 34
  • two-thirds
  • three-dozen
  • “more than 30″

The numbers above illustrate some of the ways in which different sources (news outlets, pundits, policy experts, advocates, my mom… okay, not my mom… but I bet she’d have an opinion if I asked her) are reporting the number of states that could or would (yes, that distinction is important) lose subsidies if the Supreme Court decides to take them away.

I was trying to design an infographic to illustrate how many states were at risk of losing these subsidies. I figured on about 45 minutes to create a map and a bar chart. But I couldn’t. Because there wasn’t a number. Or rather, because there were many. And each number was more loaded than the rest.

How many states would lose subsidies if the Supreme Court decides in favor of the plaintiffs in King v. Burwell? It’s like the beginning of a “How many people does it take to screw in a lightbulb…” joke. Except there’s no punchline because it’s not funny—families get to lose their health insurance over this one.

So my straightforward question took me down a long, sad rabbit hole of politics, healthcare, and the Supreme Court.

Seriously, how do you count states?

To understand why these numbers matter, you need a basic understanding of how the healthcare exchanges are structured. Bear with me for four sentences. Under the ACA, each state plus the District gets to choose whether to set up its own health insurance exchange or to let the feds do it:

  • If the state sets up its own exchange, it’s called a state-based marketplace and there are 14 (including DC)
  • If the state lets the government do it, it’s called a federally facilitated marketplace and there are 27

If the above two categories were it we’d be all squared away, just a knock-down-drag out fight over people trying to take healthcare away from low- to- middle-income people. No numbers involved. But there’s a gray area with states that have partnerships—and not everyone agrees which of the above categories these gray states could fall into if the ruling goes down.

  • Some state exchanges are called partnerships (a state-partnership marketplace and there are 7 to be painfully precise). This is where the states and the feds divvy up the work of administering the exchange and the whole thing is run off the federal website—healthcare.gov
  • There are also states that do all the heavy lifting of administration and just use healthcare.gov as a technology platform. They basically have a state-based marketplace but happen to run it off the fed website—Oregon, Nevada, New Mexico. (If you are a glutton for wonk terms this is called a federally supported state-based marketplace and there are 3 of these states).

Kaiser has a nice wonky chart that lays it all out here.

Why is the number of states so important in King v. Burwell

Now, the lawsuit in King is basically an argument over which states with which types of exchanges get to give subsidies. The plaintiffs claim that the law intended for subsidies to be provided only to states that run their own exchanges (state-based marketplaces). That requires a definition: What, exactly, is a state that runs its own exchange? And a number: How many states do that?

Then there are those three “gray area” states in play (Oregon, Nevada, New Mexico) that make counting things more complicated: Do they/don’t they meet criteria for being state-based marketplaces? Depending on who you ask, they’re either at risk (see above) or doing just fine, thank-you-very-much (see above). (I know people that would run me out of town for even asking that question. To those people I say, “Yes. They are state-based marketplaces. But not everybody agrees with you. See above.”)

If you are interested in more of an explanation of the politics behind the three “gray area,” so was I and so was Charles Gaba, who is waaaaay smarter than me. His popular ACA blog sums up a longer Politico piece here. The title is as good as the answer: In which Politico reporters Racahan Pradhan & Brett Norman FINALLY answer a question I’ve been wondering for months… (and here’s the original Politico article).

And this matters, why?

Because depending on how the lawsuit goes, which state counts as a state-run marketplace affects who gets subsidies and who doesn’t, either through how the court defines it or how the rules define it after a decision gets handed down. I don’t have the brainpower or the will to begin to cover the arguments in this blog, so I won’t attempt to do so. But it matters. Just ask the people who won’t be able to afford health insurance once they lose their subsidies.

So we have a situation where prominent sources are reporting the numbers differently, and it’s confusing, and it’s political.

These numbers are being reported very differently by, um, everyone. Not just proponents and opponents of the law. Everyone.

  • Some count 13 states plus DC as having exchanges with “state-based marketplaces”, which means that they report 37 states at risk of losing subsidies.
  • Others count the 13 states plus DC and also what I call the three other “gray area” states (Oregon, Nevada, New Mexico).
  • Others use more nebulous characterizations (see below).

Here’s a sampling of the most common reporting that you’ve likely seen already:

So why can’t a poor designer simply catch a break and just draw a map?

Well, politics, obviously. That’s part of the point of this post. To shed light on just how murky things get for the people whose job it is to make things…er… clear.

If you’ve got a day job, drawing a map can mean visualizing—not data—but the political stance of your organization.

You’re caught in the cross-fire between your craft—show what you see/know—and the advocacy goals of your organization. Or, for that matter, the politics of the issue itself independent of your organization. Things don’t always align neatly.

Oh, by the way, I did draw the map for this blog post. Please enjoy it.

Map showing how some states are counted in King v. Burwell case

It’s complicated. But there’s another catch.

And can I just say one more thing? This is a shameless plug for my own outfit. If you really are interested in what will happen to the people who will lose their subsidies (I’m totally serious), please watch this video. Regardless of what “number” we ultimately settle on, not much good will come of it if these families lose their subsidies and can’t afford to pay for health insurance. You can hear them tell you about it in their own words. It’s pretty compelling and yes, we produced it.

The “art” of compromise: Is there room for compromise in designing data graphics?

In my last post, I discussed how expectations and perceptions of designers are as important to quality data visualizations as are more conventional resources, such as time, people and money. But there is also a flip side to this–there are times when, as designers, we may be faced with a choice to compromise on how we present data. The compromises we agree to–or reject–are as important to our field as anything else. (Kudos to me for resisting the urge to title this “drawing the line in infographics.”)

A friend related to me a recent conversation in which an art director who, when presented with a bar graph of extreme values (very high and very low), asked the designer to “fudge” the size of the smaller bars. (They were visible–not hairline–but too small to comfortably fit the values inside of them. Presumably the art director wanted to nudge them up so that the numbers would fit inside of the bars.) My initial reaction was er… not favorable. I felt like a mother bear protecting her cubs (the cubs, in this tortured analogy, are the data). I may have uttered a few choice words, even.

The ethics of compromise.

But, once I calmed myself down, it occurred to me that this might be something interesting to write about. I polled a few designer and non-designer friends. What do you think, I asked. Was this a bow to art or clarity? Was it an unintentional breach of ethics or a well-intentioned attempt to make information easier to understand? Was it goal-driven or just lack of creativity? Don’t jump on the art director just yet. This isn’t about the choice that person made (that’s the subject of another post). It simply reflects the reality that, as in other professions, we’ll all be asked to make choices that, to others, may appear to be inconsequential. We need to make sure we handle these choices intentionally and carefully.

Here’s what came to mind after my conversations with other designers.

Book-binding: an invisible art

Let’s think about the book-binding trade of back in the day. The men (mostly men, anyway) who bound books hundreds of years ago were tradesmen. They had a craft which they revered. They apprenticed and, as journeymen, they traveled from place to place, learning and honing their craft to become–eventually–book-binders. This is not unlike the path that many information designers take today.

For all the painstaking zeal and meticulousness put into the binding of the book, the end result was rarely if ever examined once produced. If the thing didn’t fall apart in your hands, you were satisfied as a consumer.

I won’t bore you with the mechanics, but suffice to say that binding a book involved a lot of work, much of which was invisible to the eventual and subsequent owners. Once purchased, the book was read, perhaps the craftsmanship briefly admired, and then it was shelved or passed on, sometimes for generations (think of the family Bible). And yet, for all the painstaking zeal and meticulousness put into the binding of the book, the end result was rarely if ever examined once produced. Again, not unlike the process of visualizing data, much of the effort and care involved in sewing pages into folios, hand-stitching the spine–remained largely unseen. If the thing didn’t fall apart in your hands you were satisfied. End of story. And yet, despite this invisibility, these bookbinders pursued their craft with diligence and and care nonetheless. How well or how poorly they plied their trade was not immediately evident, as these old books often outlived their makers. They had no immediate incentive to be unduly diligent. And yet, I like to think that most of them did not cut corners. Why? I’d say it was self-respect and public recognition of the importance of their craft. Maybe I’m over romanticizing books (I do collect them).

Our craft: Are we short-order cooks or visual content experts?

My point? This is an issue of the ethics of our craft. As designers, we need to ask ourselves: are we short-order cooks or visual content experts? Are we hacks or tradesmen/women? Is data visualization a craft or only a paycheck? Is data an obstacle to be overcome or a living boundary that, with each challenge, offers us the opportunity to learn, do better, and to empower our readers by bringing information to the surface in a manner that brings with it a new understanding? And while, from the perspective of the client (or, in this case, the art director) it may not always be apparent that the accommodations they ask us to make are not wise, it is–nonetheless–our responsibility to do the right thing, and bring others along. In this way, we advance the field forward and our professionalism as well.

And that’s the crux of this post.

Whatever your intentions, what is the effect of the small compromises that you make in being precise, transparent and correct in how you present data?

The more seasoned amongst you may shake your heads and think that these things are self-evident. But to those of you who are just starting out (be it as younger designers or managers in charge of new data viz projects), this may not be something you’ve thought much about. It may not even seem like much of a big deal to you.

Making those small compromises weighs on you, wears you down and–worse–makes the next compromise all the greater in scope and easier to bear.

What is the effect of compromise on the designer and the team?

So, what happens when a designer makes those compromises? When I asked a few designers, they all had one response in common: morale and self-esteem. Here’s the thing: making that one small edit will be invisible to everyone but you. It’s not like your readers will ask you to send them your Illustrator file so that they can measure pixels before they read further. Like the bookbinder who sewed thread onto page folios, no one but you will see the guts of your files. But making those small compromises weighs on you, wears you down and–worse–makes the next compromise all the greater in scope and easier to bear. And these things add up to the slow devolution of what was once a craftsman/woman (if I may be allowed to use such an archaic term) to a hack.

And what happens when an art director suggests those compromises? Well, you risk losing the respect of seasoned members of your team, that’s obvious. Worse, you risk creating an environment that is progressively sloppy. And while no one will catch the small compromises, they sure as hell will catch the big ones. Remember the infamous Fox piechart?

Other examples of altering data

It doesn’t stop with information designers, as I’m sure you know. Another designer who Photoshops medical imagery (for example, a CT scan or slides of cancer cells) told me about a doctor who, when preparing images of slides for a research publication, asked the designer to darken some areas to make them more visible (thus allowing him to better make his case). The designer balked–these aren’t just pictures, he told the doctor–they’re data.

And if you want a more mainstream example, how about the furor over the Time cover of OJ Simpson in 1994? Or, more recently (2008), the Hillary Clinton ad which featured then Presidential candidate Obama with arguably darker skin?

What is unacceptable compromise to one might be reasonable accommodation to another.

There may not be room to make the wrong compromises, but there is always opportunity for discussion.

No one is perfect. And each of the examples that I gave leave plenty of room for discussion. As a newspaper friend recently noted, some photographers are adamant about not retouching any photos they take–including not cutting subjects out of backgrounds. Others are not as rigid. And not all of the participants in my informal poll reacted with extreme horror at the thought of slightly lengthening bars. Some merely grimaced. But all agreed that if you’re going to tread on thin ice, you’d best aware of it. Another friend points out that he noticed a disconnect between his former employer (a newspaper) and his current one (a corporation). He’s doing the same work–designing information graphics. But whilst former journalist colleagues (having their own code of ethics) would never have asked him to fudge the appearance of data, he feels that–in his current role as a designer in the corporate world–his colleagues have a lesser understanding and appreciation of what asking this might mean.

This isn’t necessarily a bad thing–handled correctly, it can present an opportunity for education. But you have to be willing to put yourself out there–a place that not everyone (perhaps less experienced designers or as employees with less seniority) is comfortable occupying.

As designers, let us be keenly aware of how the small choices we make for ourselves can add up to large consequences for our profession. I’d love to hear more from you on this. Have you been place in similar situations? How did you handle them?

The ethics of information

I don’t pretend to be an expert on any subject other than one: how to recognize a perfect pizza. That’s not false humility, it’s a candid admission. Most of what you’ll read in this blog are summaries of my learning curve in pixels–summaries built on the experience of those more patient, methodical and talented than I. Thank goodness for the interwebs.

To the wealth of information out there, I can add only a small amount of experience, most of which is gleaned from making mistakes; from not asking the right questions; from not sufficiently challenging, and thus not understanding, the premise of a project; and from occasional bouts of arrogance or foolishness. Okay, the confessional is closed. But the reason I make this point is because this is exactly where ethics and best practices come in. It’s your first line of defense against silly ideas foisted upon you by unknowing clients, editors, writers–even you.

A recent post by Alberto Cairo entitled “Infographics as Moral Acts” reminds us, yet again, that as much as we raise the bar in each and every way–via the visual arts, or through  technology, or by envisioning new ways to tell our stories through data–it doesn’t amount to much without some guiding principles. This is not a new idea, but I when I look at the proliferation of infographics I do wonder how top of mind this is for information designers (myself included). Some signs are encouraging–as some of you may remember, Visual.ly, a popular data viz sharing site, adopted a code of ethics for data visualization in February (other blogs, including Tableau, wrote about this as well, though the discussion generated little comment other than a reference to Fox News–below).

So, read this post, as well as a related article from the Harvard Nieman Watchdog Journalism Project (co-written by Mr. Cairo) which the article references, and try to make it part of your work in meaningful ways.

We’re listening to you, Alberto. But apparently, Fox News still is not.

Gas prices example from Media Matters