Choropleths and Colours Part 2

Last Thursday I wrote about the use of colour in a choropleth map from the Philadelphia Inquirer. Then on Sunday morning, I opened the door to collect the paper and saw a choropleth above the fold for the New York Times. I’ll admit my post was a bit lengthy—I’ve never been one described as short of words—but the key point was how in the Inquirer piece the designer opted to use a blue-to-red palette for what appeared to be a data set whose numbers ran in one direction. The bins described the number of weeks a house remained on the market, in other words, it could only go up as there are no negative weeks.

Compare that to this graphic from the Times.

More choropleth colours…

Here we are not looking at the Philadelphia housing market, but rather the spread of the UK/Kent variant of SARS-CoV-2, the virus that causes COVID-19. (In the states we call it the UK variant, but obviously in the UK they don’t call it the UK variant, they call it the Kent variant from the county in the UK where it first emerged.)

Specifically, the map looks at the share (percent) of the variant, technically named B.1.1.7, in the tests reported for each country. The Inquirer map had six bins, this Times map has five. The Inquirer, as I noted above, went from less than one week to over five weeks. This map divides 100% into five 20-percent bins.

Unlike the Inquirer map, however, this one keeps to one “colour”. Last week I explained why you’ll see one colour mean yellow to red like we see here.

This map makes better use of colour. It intuitively depicts increasing…virus share, if that’s a phrase, by a deepening red. The equivalent from last week’s map would have, say, 0–40% in different shades of blue. That doesn’t make any sense by default. You could create some kind of benchmark—though off the top of my head none come to mind—where you might want to split the legend into two directions, but in this default setting, one colour headed in one direction makes significant sense.

Separately, the map makes a lot of sense here, because it shows a geographic spread of the variant, rippling outward from the UK. The first significant impacts registering in the countries across the Channel and the North Sea. But within four months, the variant can be found in significant percentages across the continent.

Credit for the piece goes to Josh Holder, Allison McCann, Benjamin Mueller, and Bill Marsh.

Choropleths and Colours

In many cities through the United States, real estate represents a hot commodity. It’s not difficult to understand why, as have covered before, Americans are saving a bit more. Coupled with stay-at-home orders in a pandemic, spending that cash on a home down payment makes a lot of sense for a lot of people. But with little new construction, it’s a seller’s market.

The Philadelphia Inquirer covers that angle for the Philadelphia region and in the article, it includes a map looking at time to sell a house. And it’s that interactive map I want to look at briefly this morning.

Red vs. blue

Primarily I want to discuss the colours, as you can gather from this post’s title. We have six bins here, each indicating an amount of time in one-week intervals. So far so good. Now to the colours, we have red for homes that sell in one week or less and blue for homes that sell in five weeks or more.

Blue to red is a pretty standard choice. You will often see it in maps where you have positive growth to negative growth or something similar, I’ve used it myself on Coffeespoons a number of times, like in this map of population growth at the county level here in Pennsylvania.

In those scenarios, however, note how you have positive values and negative values. The change in colour (hue) encodes the change in numerical value, i.e. positive vs. negative. We then encode the values within that positive or negative range with lighter/darker blues and reds. Most often the darker the blue or red, the greater the value toward the end of the spectrum. For example, in Pennsylvania, the dark blue meant population growth greater than 8% and red meant population declines in excess of 8%.

As an aside you’ll note that there are no dark blue counties in that map and that’s by design. By keeping the legend symmetrical in terms of its minimum and maximum values, we can show how no counties experienced rapid population growth whilst several declined rapidly. If dark blue had meant greater than 4% growth, that angle of the story would have been absent from the map.

Back to our choropleth discussion, however. How does that fit with this map of selling times for homes in the Philadelphia region?

Note first that five weeks is a positive value. But so is one week or less. The use of the red-blue split here is not immediately intuitive. If this map were about the change or growth in how long homes sell, certainly you could see positive and negative rates and those would make sense in red and blue.

The second part to understand about a traditional red-blue choropleth is that at some point you have to switch from red to blue, a mid-point if you will. If you are talking positive/negative like in my Pennsylvania map, zero makes a whole lot of sense. Anything above zero, blue, anything below zero red.

Sometimes, you will see a third colour, maybe a grey or a purple, between that red and blue. That encodes a fuzzier split between positive and negative. Say you want to give a margin of 1%, i.e. any geographic area that has growth between +1% and -1%. That intrinsically means the bin is both positive and negative at the same time, so a neutral colour like grey or a blend of the two colours, a purple in the case of red and blue, makes a whole lot of sense.

Here we have nothing like that. Instead we jump from a light yellow two-to-three weeks to a light blue three-to-four weeks.

What about that yellow? In a spectrum of dark blue to light blue, you will see lighter blues than darker blues. But in a red spectrum, that light red becomes pinkish or salmonish depending on that exact type of red you use. (Conversation for another day.) Personal preferences will often push clients to asking a designer to “use less pink” in their maps. I can’t tell you the number of times I’ve heard that.

If that comes up, designers will often keep their blue side of the legend from the dark to light—no complaints there, or at least I’ve never heard any. But for the red side, they’ll switch to using hue or type of colour instead of dark to light red.

Not all colours are as dark as others. Blue and red can be pretty dark. Yellow, however, is a fairly light colour. Imagine if you converted the colours to greyscale, you’ll have very dark greys for blue and red, but yellow will be consistently far lighter than the other two.

The designer can use the light yellow as the light red. But to link the yellow to red, they need to move through the hues or colours between the two. There’s a whole conversation here about colour theory and pigment and light absorption vs. pixels and light emission, but let’s go back to your colours you learned in primary school (pigment and light absorption). Take your colour wheel and what sits between red and yellow? Orange.

And so if a client objects to a light pink, you’ll see a pseudo dark-to-light red spectrum that uses a dark red, a medium orange, and a light yellow. Just like we see here in this Inquirer map.

Back to the two-to-three week and three-to-four week switch, though. What’s the deal? This is my sticking point with the graphic. I am looking for the explanation of why the sudden break in colour here, but I don’t see any obvious one.

Why would you use this colour scheme where blue and red diverge around a non-zero value? Let’s say the average home in the region sells in three weeks, any of the zip codes in red are selling faster than average, hot markets, and those taking longer than average are in blue, cold markets. Maybe it’s the current average, however. What if it were the average last year? Or the national average? These all serve as benchmarks for the presented data and provide valuable context to understand the market.

Unfortunately it’s not clear what, if any, benchmarks the divergence point in this map reflects. And if there is no reason to change colours mid-legend, with only six bins, a designer could find a single colour, a blue or purple for example, and then provide five additional lighter/darker shades of that to indicate increasing/decreasing levels of speed at which homes sell.

Overall, I left this piece a wee bit confused. The general trend of regional differences in how quickly homes are selling? I get that. But because there’s a non-logical break between red and blue here—or at least one I fail to see in the graphic—this map would work almost as well if each bin were a separate colour entirely, using ROYGBIV as a base for example.

Credit for the piece goes to John Duchneskie.

Discontinuous Lead Bars

Last week the Guardian published an article about drinking water pollution across the United States. Overall, it was a nicely done piece and the graphics within segmented the longer text into discrete sections. Each unit looks similar:


The left focuses on a definition and provides contextual information. It includes small illustrations of the mechanisms by which the pollutant enters the water system. To the right is a chart showing the levels of the contamination detected in the 120 tests the Guardian (and its partner Consumer Reports) conducted.

In almost all of the charts, we see the maximum depicted on the y-axis. And the bars are coloured if that observation station exceeds the health and safety limits. (The limit is represented by the dotted line.)

But towards the end of the piece we get to lead, a particularly problematic pollutant. There is no safe level of lead contamination. But how the piece handles the lead chart leaves a bit to be desired.

But how bad is it, really?

The first thing is colour, but that’s okay. Everything is red, but again, there is no safe level of lead so everything is over the limit. But look at the y-axis. That little black line at the top indicates a discontinuity in the lines, in other words the values for those three observations are literally off the chart.

But does that work?

First, this kind of thing happens all the time. If you ever have to work with data on either China or India, you’ll often find those two nations, due to their sheer demographic size, skew datasets that involve people. But in these kind of situations, how do we handle off the charts data points?

There is a value to including those points. It can show how extreme of an outlier those observations truly are. In other words, it can help with data transparency, i.e. you’re not trying to hide data points that don’t fit the narrative with which you’re working.

In this piece, it’s never explicitly stated what the largest value in the data set is, but I interpret it as being 5.8. So what happens if we make a quick chart showing a value of 6 (because it’s easier than 5.8)? I added a blue bar to distinguish it from the the rest of the chart.

It’s pretty bad.

You can see that including the data point drastically changes how the chart looks. The number falls well outside the graphic, but it also shows just how dangerously high that one observation truly is.

But if you say, well yeah, but that falls outside the box allowed by the webpage, you’re correct. There are ways it could be handled to sit outside the “box”, but that would require some extra clever bits. And this isn’t a print layout where it’s much easier to play with placement. So what happens when we resize that graphic to fit within its container?

And resized

You can see that All the other bars become quite small. And this is probably why the designers chose to break the chart in the first place. But as we’ve established, in doing so they’ve minimised the danger of those few off-the-charts sites as well as left off context that shows how for the vast majority of sites, the situation is not nearly as dire—though, again, no lead is good lead.

What else could have been done? If maintaining the height of the less affected bars was paramount, the designers had a few other options they could have used. First, you could exclude those observations and perhaps put a line below the 118 text that says “for three sites, the data was off the charts and we’ve excluded them from the set below.”

I have used that approach in the past, but I use it with great reluctance. You are removing important outliers from the data set and the set is not complete without them. After all, if you are looking to use this data set to inform a policy choice such as, which communities should receive emergency funding to reduce lead levels, I’d want to start with the city in blue. Sure, I would like everyone to get money, but we’d have to prioritise resources.

I think the best compromise here would have actually been a small tweak to the original. Above the three bars that are broken (or perhaps to the right with some labelling), label the discontinuous data points to provide clearer context to the vast majority of the sites, which are below 0.5 ppb.

As easy as ABC

This preserves the ability to easily compare the lower level observations, but provides important context of where they sit within the overall data set by maintaining the upper limits of the worst offenders.

Credit for the piece goes to the Guardian’s graphics department.

Too Much Horsing Around

Last week the Philadelphia Inquirer published an investigation of the staggering number of horse deaths in Pennsylvania’s race track facilities. I found the article fascinating, but admittedly at a point or two a wee bit squeamish when the author described how horses essentially die. Then about halfway through the article I ran into the first of two graphics looking at the data.

Seeing red…

The first is pretty simple, a timeline of deaths over the course of one year, 2019. Overall it works, you can clearly see clusters of racing deaths, but that those clusters spread across the year. When I sat with the graphic for a moment, however, a few things began to stick out at me. The first was a distracting vibration in the background. Not the alternating beige and blue of the months, but if you look closely you’ll see tightly spaced lines within the colour fields: presumably the days of the month for aligning the deaths.

On a large enough graphic it makes all the sense to tick off sub-monthly increments, but in this space I would have probably opted to show only the months. Maybe weeks could have worked, as that approach may have reinforced the statistic about a horse dying every six days on average.

The second point is the black stroke or outline of each dot. Here the designer faces a challenging constraint. Essentially, the smaller the dot (or the symbol) the brighter the colour. In a rich, blood red colour you have a dark heavier colour. Compare that to say a stop sign that is bright red. It has a lighter feel. The blood red colour, in a given space, has let’s say an amount of black ink or pixels—I’m simplifying here—mixed in with the red. But in a large area, there’s enough red ink or pixels to still be clearly blood red. The stop sign red has no other colours but red. And in large areas, it can be an eye-stabbing amount of red—precisely why it’s likely so useful for, you know, stop signs.

But at the small scale of these very small dots, you still proportionally have the same amount of red and black ink, but with fewer and fewer amounts, the eye can begin to experience difficulty in truly reading the colour for what it is. For example, in an area of say 49 pixels (7×7), while the ratio of red to black may be consistent, you still only have a total of 49 pixels with which to convey “red” to the reader. Consequently, in smaller spaces, you may find that designers sometimes opt for brighter colours, a la stop sign red, than they would in larger fields of colour.

Here we have a nice use of brighter red, green, and yellow. (I will quickly add that the choice of red and green can be problematic for colour blindness, but I don’t want to revisit that here.) But to provide better separation between those small, circle sized fields of colour a border probably helps. A thin black line, or stroke, makes sense. But the black is darker than the colours themselves, thus it can draw more attention than the colour fill. And that begins to happen here. I wonder if a thin white stroke may have been less distracting and placed more emphasis on the fill colours.

As I said, overall a really nice if not sobering graphic in an important but disturbing article. I think a few small tweaks could really bring the graphic over the finish line. Pun fully intended. Sorry, not sorry.

Credit for the pieces goes to John Duchneskie.

Making America Save Again

For years, one issue with the American economy had been that we did not save enough. It’s understandable, as it’s hard to keep up with the image of the carefree American without profligate spending. But that’s also not great long-term. But thanks to Covid-19, we’ve now swung to the other side of the spectrum: Americans may be saving too much.

Saying that sounds callous to the devastation the pandemic has wrought upon large swathes of the economy. But it’s true in the aggregate as this New York Times piece explains. In particular, the authors highlight one example. Consider a corporate CEO who earned a $100,000 bonus for keeping the company he runs afloat during the recession. He adds $100k to the aggregate American income. But at a restaurant shuttered by the pandemic, owners lay off a hostess, a server, a bartender, and a dishwasher, each earning $25,000. Their collective lost income is $100,000 and so balances out that one CEO. And as CEOs are more able to work remotely than servers, it’s not hard to see how the upper-income earning cohorts of the economy have done well. In human-terms, four unemployed service industry people is terrible. But statistically, it’s a wash. Once we understand that, it makes the piece sensible.

It uses decomposition charts, basically stacked bar charts broken apart, to show what constitutes the two sides of the American household budget: earning and spending. I’ve taken a screenshot of the spending side of the ledger.

This is the aggregate, I’d be curious how this relates to you, my readers.

We see that starting from the baseline, the solid line, American households spent more money this year on durable goods. A dotted line then carries that adjusted baseline to the right for the next component of the ledger: nondurable goods. We spent more on those too, so the baseline moves up. The designers annotated the graphic, adding descriptions of what each bar represents in a casual, lighthearted tone. I’ve definitely been cooking for myself a lot more.

Here I wish we had some more traditional charting elements, e.g. axis lines and labels. Now this piece is published under the Upshot, a more conversational and less formal brand than the Times as a whole. That probably explains the casual annotations. But I think some basic axis labels, e.g. spending more vs. spending less, could add some context without the need for the annotations.

Where the piece might lose people is what happens after durable goods. Americans stopped spending on services, a decline of over half a trillion dollars. That’s a lot of money. And so the adjusted baseline shifts to well below where we started. Add on savings from things like interest rates (Jay Powell is the chair of the Federal Reserve, for whose Philadelphia bank I work in full disclosure) and Americans have spent more than half a trillion dollars less. And as the article explains, we’ve also saved an enormous amount, to the tune of $1 trillion. Add it together and you’ve got America saving $1.5 trillion in 2020.

That money has to go somewhere. And you can see where some of it went when you look at surging prices in GameStop. Longer term, when the pandemic begins to end, we are going to have a pent up demand from people who have had their lives on hold for a year or more. And if there is insufficient supply for whatever’s in demand, prices will rise and we could see a sharp jump in inflation. But that’s a post for another day.

Back to this graphic, as a statistical graphic, it works. But without axis labels and data definitions, barely so. However, I think it’s meant to be more casual and illustrative than data-driven. If I look at this piece through that lens, I do think it works.

Credit for the piece goes to Neil Irwin and Weiyi Cai.

Farewell, Cardboard Cutouts

In 2020, baseball did not permit fans to attend regular season matches. (They changed this for the playoffs.) Instead, many stadiums opted for cardboard cutouts: fans often paid a fee and submitted a picture that the team printed on cardboard cutouts. Like so many things we will say about 2020, it was surreal.

But in Philadelphia at least, cardboard cutouts are out, and human fans are in. The state government in Harrisburg and the city government will allow 20% capacity at outdoor stadiums and 15% for indoor stadiums.

The Philadelphia Inquirer created a small graphic for its homepage to capture this news.

I cannot wait to safely attend a live match. C’mon, vaccines.

I intentionally included other site elements in the cropping to show how the graphic fits into the broader site. The extra white space around the image helps focus attention on the datagraphic over the numerous photographic elements for each article. Clicking on other tabs in the section brings up full-component-width graphics.

To the graphic itself.

Still can’t wait…

My guess would be this was a quick turnaround piece. There are a few things going on here. The first and most obvious one, the squares as spectators. Now I confess this confused me at first. I was not entirely certain what the coloured squares meant; they mean in-person attendees. Was this supposed to be an overall stadium? Or was it a representative seating section?

The quick turnaround becomes important, because this is probably how I would have first conceptualised the graphic. But, with more time, I may have attempted to incorporate the shape of the playing field, be it a baseball diamond or basketball court, or hockey rink—I know all the sports terms!—and surrounded them with shapes representing a certain number of spectators. Squares might not work in that case because of the curves. Circles? Hexagons? Regardless of the shape, the filling of occupied seats would be the same as here, but it would perhaps be clearer to some readers, i.e. me.

Second, we get to the table below the graphics. Here we have a subtle design decision. Note that here the designer greyed out the normal capacity figures. The new figures at that 20% and 15% rates are what appear in black bold text. My usual instinct is to use typographic weight, regular vs. bold, in these situations. But the grey here works equally well.

Third, and this also involves the table, we have the first game data. We talked about the comparison of the capacity and permitted attendance. But I wonder, did the date of the first game with fans needed to be displayed in the same way as the permitted attendance? Because the news isn’t the dates of the first games—at least not as I read the news—but the numbers of attendees. And because of that, maybe I would have reduced the size of the type for the date of the first game. Or, conversely, set the type for the new attendance in a larger point size.

Overall, I enjoyed seeing this news presented visually, even if I was left confused.

Credit for the piece goes to John Duchneskie.

Lead Pie

This past weekend, I read an article in Politico discussing parents’ outrage over levels of lead and other toxic metals in baby food. The story focuses on a Congressional report into the matter, but that ties back into an EPA study from 2017 that investigated lead contamination. Specifically the article’s author notes “a chart that was buried in supplemental material”. Buried chart? Well I went off to investigate.

And I found all the charts. But I wanted to focus on one. I am not entirely clear what it means: Percent contribution by pathway adjusted for bioavailability of each media for NHEXAS Region 5 study. I get that it’s looking at channels of intake, but it’s unclear if this is lead or some other contaminant. Is this for all people? Or a sub-section of the population as other charts in that supplemental material pack are?

So I made a graphic where I compared the original to two alternate versions.

Now, the editorial focus of the article is on baby food, which is not the apparent focus of the study (unless it is couched in academic/technical terms). But what’s worth noting is that the pale yellow recedes into the background as the burgundy dominates the graphic.

If graphics are done well, they should show clear visual relationships, they do not need to label specific datapoints unless through a progressive disclosure of information. But if you are going to label everything, I would want to make certain that in the case of that same burgundy slice, we have sufficient contrast to read the 17% label.

Pie charts are not good at allowing people to compare slices. So the pie chart as the format here is not a great place to start, but as you can see in my Option 2, if you are going to choose a pie chart form, there are ways of making it more legible. Namely, do not make it three-dimensional.

Here the foreground receives prominence over the background, which may be receding and visually shrinking into the background. And as the point of a chart is to make visual comparisons, if we cannot compare like for like, it’s not ideal.

Also, we have the thickness of the pie chart. That vertical heights adds yellow to the slice of the pie we see in front. Casually, that makes the yellow slice appear even larger than it already is from the three-dimensional foreshortening.

Option 2 presents this as a stripped down pie chart. Make it flat. I used one colour with tints of one purple. I used the 100% to highlight the dietary intake channel, because of the Politico article’s focus.

But really, Option 1 is the improvement here. Comparing the smaller slices is easier here as the eye simply moves vertically down the graphic. We are also able to add axis lines that provide a context for where those values fall, between 0 and 10 for Water intake, and just over 10 for Air. Somewhere between 15 and 20 for Soil and dust ingestion.

Finally, that legend. We don’t want the reader to have to strain to identify what slice is what. Why is the legend in a box? Why is it so far away from the pie? In both my options I closely and visually link the labels to the slices/bars they represent. That makes it easier for the reader to know what they are looking at when they are looking at it.

The moral of the story, people, don’t use three-dimensional pie charts.

Credit for the original version goes to the EPA. Credit for the alternate versions is mine.

Another Look at 500,000

Yesterday we looked at how the New York Times covered the deaths of 500,000 Americans due to Covid-19. But I also read another article, this by the BBC, that attempted to capture the scale of the tragedy.

Instead of looking at the deaths in a timeline, the BBC approached it from a cumulative impact, i.e. 500,000 dead all in one go. To do this, they started with an illustration of 1,000 people. Then they zoomed out and showed how that group of 1,000 fit into a broader picture of 500,000.

We’re going to take a look at this in reverse, starting with the 500,000.

Half a statistic.

I think this part of the graphic works well. There’s just enough resolution to see individual pixels in the smaller squares, connecting us to the people. And of course the number 500 stacks nicely.

My quibble here might be whether the text overlay masks 8,000 people. Initially, I thought the design was akin to hollow square, but when I looked closer I could see the faint grey shapes of the boxes behind a white overlay. Perhaps it could be a bit clearer if the text fell at the end of all the boxes?

But overall, this part works well. So now let’s look at the top.

1,000 tragedies

This is where I have some issues.

When I first saw this, my eyes immediately went to the visual patterns. On the left and right there are rivers or columns of what look like guys in white t-shirts. Of course, once I focused on those, I saw other repeated patterns, the guy in the black jacket with his arms bent out, hands on his hips. The person in the wheelchair occupies a different amount of area and has a distinct shape and so that stood out too.

Upon even closer inspection, I noticed the pattern began to repeat itself. Every other line repeated itself and with the wheelchair person it was easy to see the images were sometimes just flipped to look different.

Now, allow me to let you in on a secret, unless you gave a designer a budget of infinite time, they wouldn’t illustrate 1,000 actual people to fill this box. We don’t have time for that. And I’ll also admit that not all designers are good illustrators—myself first and foremost. A good design team for an organisation that uses illustration should have either a full-time illustrator, or a designer who can capably illustrate things.

But this gets to my problem with the graphic. I normally can distance myself from reading a piece to critiquing it. But here, I immediately fixated on the illustrations, which is not a good sign.

There are three things I think that could have been done. The first two are relatively simple fixes whilst the third is a bit grander in scope.

First, I wonder if a little more time could have been spent with the illustrations. For one, white t-shirt guy, I don’t see his illustration reused, so why not change the colour of his t-shirt. Maybe in some instances make it purple, or orange, or some other colour. I think re-colouring the outfits of the people could actually solve this problem a good bit.

But second, if the patterns still appear visible to readers, mix it up a bit. I understand the lack of desire to spend time creating an individualised row for each row. Crafting each row person by person probably is out of the time requirements—though maybe the people above the designer(s) should know that content takes time to create. So what about repeating smaller blocks? I counted 20 rows, which means there should be 50 people per row. Make each row about ten blocks, and have several different blocks from which you can choose. Ideally, you have more blocks than you need per row, so not all figures are repeated, but if constrained, just make sure that no two rows have the same alignment of blocks.

Thirdly, and here’s the one that would really have required more time for the designer to do their job, make the illustrations meaningful. In a broad sense, we do have some statistics on the deaths in the United States. According to the CDC, 63% of deaths have been by white non-Hispanics, 15% by Black non-Hispanics, and 12% by Hispanic/Latino, 4% by Asian Americans, 1% by Native Americans, 0.3% by Hawaiian and Pacific Islander, and 4% by multiple non-Hispanic. Using those numbers, we would need 630 obviously white illustrations, 150 obviously Black, and so on.

If the designer had infinite time, the illustrations could also be made to try and capture age as well. Older people have been hit harder by this pandemic, and the illustrations could skew to cover that cohort. In other words, few young people. According to the CDC, fewer than 5% of deaths have been by people aged under 40. In other words, no baby illustrations needed.

That’s not to say babies haven’t died—87 deaths of people between 0 and 4 have been reported—but that when creating a representative average, they can be omitted, because that’s less than 0.1%, or not even 1 out of 1000.

To reiterate though, that third concept would take time to properly execute. And it would also require the skills to execute it properly. And I am no illustrator, so could I draw enough representative people to fake 1,000? Sure, but time and money.

The first two options are probably the most effective given I’d bet this was a piece thought up with little time to spare.

Credit for the piece goes to the BBC graphics team.

Dove vs. Hawk

Earlier, I saw these two graphics floating around the Twitter. They each come from a major financial institution and attempt to place the voting (and non-voting members) of the Federal Open Market Committee (FOMC) on a spectrum of doves to hawks or slightly less dovish. The FOMC, part of the Federal Reserve system, sets interest rates for the US economy. Now, I’m being super simplistic here, but it’s broadly true. I should add, full disclosure, I presently work for the Federal Reserve Bank of Philadelphia.

The first graphic is from JPMorgan and plots in one-colour all the voting and non-voting members on a single axis from very dovish to somewhat less dovish. Thin black lines point to evenly spaced points on the axis and people are listed at each interval.

It’s a fairly simple approach, but effective. Nothing revolutionary here. What I find a bit odd is the line underneath the centre tick. What prompts that group to have what I’ll call a summary bar? Is it because Jay Powell, the chair of the Federal Reserve, is placed within that group? It’s a bit unclear.

Now keep in mind the classifications here, very dovish and somewhat less dovish, as we compare JPMorgan’s graphic to that of Bank of America.

The first thing that strikes me is the use of colour. Here we have a fairly straightforward divergent spectrum of red to blue. Along with other design elements, like typographic scale and contrast for the header, subhead, and labels, this piece strikes me as better designed and more polished.

But I still have questions.

Here we have dovish to hawkish. At the hawkish extreme, we have Esther George of Kansas City and Robert Kaplan of Dallas. In JPMorgan’s chart, both are grouped together as somewhat less dovish. But with Bank of America, they are decidedly hawkish. (Although with nine intervals, the Bank of America graphic has a bit more granularity than JPMorgan’s.)

So the biggest question, unfortunately left unanswered by each graphic, is what defines hawkish and somewhat less dovish? Just by words, they sound not at all alike. But both companies clearly place both individuals at the same end of the spectrum.

Part of the issue stems from the divergence point between red and blue. For most spectra of this type, that would be the demarcation between a committee member who is a dove or a hawk. But we have no similar separation for JPMorgan.

There is, however, one design element for Bank of America’s piece that I really like. My explanation of the FOMC at the top was a bit simplistic. Not every regional Federal Reserve president gets to vote every year. They rotate each year except for New York. These presidents get to vote alongside those on the Board of Governors.

In the graphic, note that everybody above the axis label is a member of the Board, i.e. they get to vote every year until their term expires. Below the axis we have the rotation schedule. Each line represents a bank president who can vote in a particular year. For example, the Philadelphia president, Patrick Harker, was a voting member on the committee in 2020, but falls off in 2021 and will not return to 2023. The Bank of America graphic captures this for each president very well.

I am a bit confused as to why some members, i.e. Kaplan and John Williams of New York, appear to sit between lines. I am unaware of any reasons why they would be between years.

Overall, I prefer the Bank of America piece. It more clearly presents the rotation element of the voting members of the FOMC. Yes, it has colours, but I’m confused as to why the demarcation between doves and hawks happens where it does. And why JPMorgan doesn’t describe anyone as a hawk. So while I prefer it, I think it could still use some additional information or context to make it clearer to readers.

Credit for the JPMorgan piece goes to a designer at JPMorgan.

Credit for the Bank of America piece goes to a Bank of American Global Research designer.

Needle Time

Yesterday was maybe the last election day for the 2020 US General Election. (There are still a few US House seats yet to be called, most notably a contested race in upstate New York.) These were a pair of runoff elections in Georgia for the state’s two US Senate seats (one for a full, six-year term, the other to finish out the final two years of a retiring senator).

I spent most of the night eating pizza and tracking results. One thing that I keep tabs on (in the sense of open tabs in the browser) is the New York Times needle forecast. It has its problems, but I wanted to highlight something I think was new last night. Or, if it wasn’t, I didn’t notice it back in November.

Below the needle was a simple table of results.

The needle speaks

In the past, the needle was a bit opaque and it consumed data and spat out forecasts without users having a sense of what was driving those forecasts. Back in November, there were a few instances where states published incorrect data—that they later fixed—and when the needle consumed it, the needle forecast incorrect results.

But now we have a clear record of what data the forecast consumed in the table below the needles. It’s fairly straightforward as tables go. But tables don’t have to be sexy to be clear and effective.

The table lists the time when the data was added, the number of votes added, the type of vote added, and then the actual data vs. what was expected. And ultimately how that changed the needle. This goes a long way towards data transparency.

Simple colour use, bright blues and reds, show when the result/data favoured the Republican or Democrat. Thin, light strokes instead of heavy black lines for rows and columns place the visual emphasis on the data. And smaller type for the timestamp places the less important data at a lower level of importance.

It’s just very well done.

Credit for the piece goes to Michael Andre, Aliza Aufrichtig, Matthew Bloch, Andrew Chavez, Nate Cohn, Matthew Conlen, Annie Daniel, Asmaa Elkeurti, Andrew Fischer, Will Houp, Josh Katz, Aaron Krolik, Jasmine C. Lee, Rebecca Lieberman, Jaymin Patel, Charlie Smart, Ben Smithgall, Umi Syam, Miles Watkins and Isaac White.