I missed last week’s posting on an update to Covid-19. Two weeks on from the last post, things in the states of Pennsylvania, New Jersey, Delaware, Virginia, and Illinois continue to improve, albeit with a few fits and starts. But the downward trend nonetheless can be seen in the new cases charts.
Consider that in the charts from two weeks ago, we saw downward slopes, but a look at the charts in the two weeks hence shows some blips.
Another thing to keep in mind is that a major snowstorm disrupted testing and vaccinating operations in the northeastern states of Pennsylvania, New Jersey, and Delaware. The storm, which also hit northern Illinois and Virginia, also likely impacted those states but to lesser degrees.
New cases curves for PA, NJ, DE, VA, & IL.
That means the downward trends in new cases could be slightly exaggerated in those states. Consequently, rebounds next week should be taken with a grain of salt. Indeed, Sunday’s data releases from the tri-state area were greater than we might normally see with weekend data.
When we at deaths, however, we see a more muddled picture.
Death curves in PA, NJ, DE, VA, & IL.
In states like Delaware and Virginia, the average death rate is now higher than it was two weeks ago. In New Jersey, the rate is down slightly, but after two weeks of it being largely up and so all in all, largely a wash. Instead, it’s only in Pennsylvania and Illinois where we any real improvements in the average death rate. Both states are down and look to continue heading down.
Finally, we look at vaccinations and the percent of state populations that have been fully vaccinated.
The fully vaccinated percentage of the populations of PA, VA, & IL.
Two weeks ago, Pennsylvania and Illinois had just reached 1%. Neither New Jersey nor Delaware is reporting similar data, so both those states remain outside our consideration set. But, all three remaining states—Pennsylvania, Virginia, and Illinois—are now over 2%. Pennsylvania reports at least 2.5%—the city of Philadelphia reports separately from the statewide Department of Health, but does not update its figures at the weekend and so is likely higher. Both Virginia and Illinois have reached 2.3% full vaccination.
Last week we saw some indications that the recent surge was beginning to ebb in Pennsylvania, Delaware, and Illinois with the same in New Jersey, but to a slight degree less so. Only Virginia presented us with data that showed its surge continuing unabated.
So this week we have some generally good news to look at.
New case curves for PA, NJ, DE, VA, & IL.
The drop in Pennsylvania, New Jersey, and Illinois appears real and sustained. Even in Virginia, we are beginning to see some signs of a decline in new cases—albeit it after a week of record reports of new cases.
Of course we should also mention that even though we are seeing declines in new cases, in no state are we close to approach low levels of community spread. Things are still bad out there, but they have gone from catastrophic spread to merely a disaster. Illinois is probably the closest to reaching summer-like levels of viral spread.
Deaths, however, because they lag behind new cases, are just now beginning to show signs of ebbing.
Death curves in PA, NJ, DE, VA & IL.
If last week’s pattern with new cases was that we were seeing positive trends in four states, we can say this week we are seeing positive trends in deaths for the same four states. Virginia is, again, the outlier.
Though I would be remiss if I noted that the declines in deaths is not nearly as pronounced as in new cases. In Pennsylvania, the seven-day trend for new deaths has appeared to have crested. But in New Jersey, recent days have suggested the decline may not be as steady. Only in Illinois are we really seeing a sustained downward trend in deaths.
And Virginia just Saturday saw its seven-day trend reach another new record, over 50 deaths per day.
But what about vaccinations?
Firstly, we still only have data for the three states of Pennsylvania, Virginia, and Illinois. Secondly, keep in mind that I am looking only at people reported fully vaccinated, i.e. they have had both their shots—both Pfizer’s and Moderna’s vaccines require two shots.
Vaccination curves in PA, VA, & IL.
There’s not a lot to report on yet, other than that both Pennsylvania and Illinois reached the 1% threshold. I think that for most people, however, that you can begin to see their respective lines easing off the 0% baseline. Virginia lags behind those two states, however, with just 0.5% of its population reported as fully vaccinated.
I’m curious to see if I cannot find some additional/alternative data sources for New Jersey and Delaware next weekend. I don’t love the idea of mixing data sources, but after a few weeks, we haven’t really seen any improvements to the data sharing from those states.
That said, I should also note that the new US administration has identified data transparency as an issue—or the lack thereof—in the current vaccination programme and is working to develop national and state-level dashboards to inform the public.
Last week we saw that in the weeks after Christmas, new cases and deaths rebounded in the five states of Pennsylvania, New Jersey, Delaware, Virginia, and Illinois. The question was how bad would things continue to get? Would these rebounds sustain themselves?
A week later we can see a glimmer of good news in that with new cases, these rebounds appear to have crested and are now ebbing back down. At least in four states.
In Virginia, unfortunately, we see that new cases continue to climb with a new record of nearly 10,000 cases reported late last week. More broadly, this is the dilemma that confronts the United States. We have states like Pennsylvania, Delaware, and Illinois where we are bringing the virus back to heel. But in other states like Virginia, things continue to get worse.
New Jersey is somewhere in the middle. It appears to just possibly be cresting with its average actually ticking higher the last few days despite falling daily new cases. We will need to see how the Garden State plays out over the course of this week.
When we look at deaths, we continue to see the grim numbers pile up.
Deaths, of course, lag new cases by 2–4 weeks, sometimes as many as six or longer. In most of our five states, the average rate of deaths appears to be cresting or peaking. In Pennsylvania the curve may have peaked. In Delaware, we have seen a plateau and in Illinois we see the best news of a resumed decline.
In both New Jersey and Virginia, however, we see deaths continuing to climb, and in some cases by significant amounts.
If cases really have peaked in some of these states over the last week, we may expect deaths to continue to rise over the course of this week before beginning to fall again.
I also want to add two new graphics today. I have been trying to figure out how to cover the vaccination programme of the five states. Unfortunately, they do not all report the same data in the same way.
The graphic that perhaps makes the most sense is the one that looks the emptiest at the moment.
In order to resume “normal” lives, we need to achieve herd immunity. When we reach that level, the virus starves of new hosts and dies out. Broadly speaking, we have two ways of achieving herd immunity.
Option 1, let the virus run rampant and takes its course through the population. The benefit is that society remains open and people can return to cafes, pubs, shops, and museums. The cost is that millions get sick and hundreds of thousands die. Sadly, this is the route taken by Sweden and, unofficially, the United States.
Option 2, vaccinate the population. The benefit here is that millions do not get sick and hundreds of thousands do not die. The cost is that in order to wait for a vaccine and vaccination we would need to close cafes, pubs, shops, and museums.
The reality is that we chose something between the two. In the initial months, after we (belatedly) recognised the threat of the virus, we shut down our economies and stayed home. We chose option 2. You can see in the state charts above how that quickly helped us curb the spread of new infections.
Unfortunately, then the Trump administration chose to follow option 1 and encouraged states to “reopen” their economies. And because we never got the virus fully under control, we sowed the seeds for the explosive growth this autumn and winter.
But the vaccines are now here and the best bet is to vaccinate the population. How many people do we need to vaccinate? The exact number depends upon the infectiousness of the virus. Measles, one of the most infectious viruses out there, requires near 100% vaccination rates to achieve herd immunity. Thankfully, this coronavirus is not as infectious as measles. Early estimates placed the range at 60–70%. But lately, some epidemiologists have indicated the true number may be higher. Dr. Fauci of the National Institutes of Health (NIH) has said the true number is likely 70–85%.
This is why the new strains of the coronavirus we have identified in South Africa and the United Kingdom worry folks. Both appear to be more transmissible than earlier strains. Neither strain appears to be more lethal in its own right—although more cases means more people will die—but this increased infectiousness could mean we need an ever higher level of herd immunity, which means more vaccinations. And we’re already seeing the anti-vaccination support rising to somewhere in the range of 15-20%, just the threshold we could perhaps tolerate with the higher herd immunity range.
So what about the chart?
As we begin vaccinations, some states are reporting the numbers of people in their state that have been fully vaccinated against the coronavirus. I plot those numbers here. Pennsylvania, Virginia, and Illinois do so. Unfortunately, neither New Jersey nor Delaware does. I only have one data point recorded for Virginia and Illinois, and so they are not plotted yet, but both fall below the level of Pennsylvania, which has reported 0.50% of its population fully vaccinated. I have added a bar to show the range of estimated herd immunity we need.
And that gets us to the second new chart, the number of total doses administered per day.
Functionally this resembles the usual two charts. We track the number of doses administered daily and then plot their seven-day average to smooth out any day-to-day blips. Of course this means almost the opposite of those two charts as we are tracking the progress of people who will be immune from the virus.
The catch is that with the current vaccines we need two shots for a full course of treatment and not all states break the data down with that level of granularity. Again, we are looking at Delaware and New Jersey as they provide only the total number of doses administered. Now that’s still helpful, but it doesn’t give us the most accurate picture of what is happening with vaccinations.
But in order to make things comparable across five states, I have decided to use that broader, total doses administered metric for Pennsylvania, Virginia, and Illinois. (Virginia and Illinois provide another headache in that it reports the daily number of people fully vaccinated, but does not break down the number of full vaccination doses.)
So what is this second chart showing us?
Well, we are seeing a slow, nearly steady growth in the number of vaccines administered. The problem is that we need to see steep, nearly exponential line charts here if we want to have any hope of returning to “normal” anytime soon. Reporting tells us that the federal government’s approach to the logistics of vaccine distribution has been…not great. (Although at this point, perhaps that should not surprise us.)
Until we see these second charts begin to show more exponential growth, the first charts of the number of people fully vaccinated will be far below that herd immunity threshold we need to see.
Covering the vaccines in addition to the virus is a bit more work, but I’m going to try and cover them both over the next several months as I have with the outbreak itself.
I meant to publish this yesterday, but this piece also offers a reminder that the hardest part of a data-driven story is usually finding the data. I was unable to find a single source of data for all the numbers I needed by the time I switched on for work. And so this had to wait until last night when I found what I needed.
And of course upon waking up this morning I found a few new articles with the data and more recent figures.
Since 2016, Trump has made building a great, big, beautiful wall on the US-Mexican border his signature policy. Of course, most illegal immigrants cross the border legally at checkpoints and normal ports of entry. A significant number are people who overstay the limits on their visas. So the efficacy of a great, big, beautiful wall is really not that great.
He also claimed that he would make Mexico pay for it.
So as he prepares to leave office, Trump this week is going on something of a victory tour and touting up his administration’s successes. The first stop? Alamo, Texas to highlight his wall.
Let’s look at that wall and how much the administration has accomplished.
For context, the US border with Mexico is nearly 2000 miles long. As of 18 December, the administration had built 452 miles, less than a quarter of the border’s total length.
Crucially, most of that construction merely replaced sections of existing wall and fence scheduled for replacement. The total amount of new wall built, as of 18 December, totals about 40 miles.
The cost of that 452 miles? More than $15 billion.
The last time we checked in on Covid-19 in the states of Pennsylvania, New Jersey, Delaware, Virginia, and Illinois, things were peaking across the five states. As I said then:
If you look at the very tippy tip top of the curves in the other four states, we might just be seeing an inflection point.
And in the month since, my highly scientific term of “tippy point” appears to have been correct. New cases did begin to drop and by the start of the Christmas holiday we began to see real improvement. I should point out that deaths continued to rise, however, but we should expect that because deaths lag new cases by sometimes as many as four to six weeks.
So how are things now, a month hence?
The new case curves for PA, NJ, DE, VA, & IL.
Well as you can see with new cases, not great and getting worse. Pennsylvania, New Jersey, Delaware, and Illinois all bottomed out prior to the holidays, and since then have been rising. It speaks to a surge in new cases likely caused by gatherings centred on the holidays.
The good news—if you can call it that—is that in Pennsylvania and Illinois, whilst cases rebound, they have not yet reached their mid-December peak in Pennsylvania and mid-November peak in Illinois. It’s worth pointing out that Chicago and separately Illinois instituted lockdowns earlier than the other four states prior to the holidays. That may account for the more dramatic reduction in those states.
The bad news is that in New Jersey and Delaware, the rebounds have now surpassed the peaks we saw in mid-December and cases continue to climb with new daily records pointing towards escalation of new cases in those states.
But the really bad news is in Virginia, where the inflection point was there—note the little mini “W” at the top of the chart—but that new cases declined neither significantly long nor in significant numbers such that there was no real holiday decline. Instead, at best we could describe it as numbers paused for two weeks before resuming their upward trends.
How about deaths?
Death curves in PA, NJ, DE, VA, & IL.
Again, fairly grim news here. A month ago we were talking about rising rates of deaths in all but Illinois. And in fact, Illinois is the only state where the death rate is significantly lower than what it was in mid-December.
In New Jersey and Virginia, we see two states where the rising death rate perhaps slowed, but it never really entered into decline. Pennsylvania and Delaware offer perhaps static death rates. In fact, Pennsylvania just yesterday surpassed its mid-December peak level.
But keep in mind that deaths lag new cases by somewhere between two to four weeks, sometimes longer. What this means is that with new cases now rebounding and in fact surpassing their peaks from a month ago, we can expect that the end of January and beginning of February could be particularly deadly.
The situation is dire in the United States and things are going to get worse before they get better.
So as begin to head into winter, where are we at with the spread of Covid-19 in the five states of Pennsylvania, New Jersey, Delaware, Virginia, and Illinois?
Nowhere good. Let’s take a look.
New cases curves for PA, NJ, DE, VA, & IL.
If you recall where we were at last week, also not great but better, cases had resumed rising post-Thanksgiving across the board. The data from yesterday indicates that cases have continued to rise everywhere but Illinois, which initiated a lockdown earlier than the other states we cover.
But Philadelphia did eventually institute a lockdown and eventually the rest of the Commonwealth followed, and similar measures—none of course as significant as those from the spring—were enacted in other states.
If you look at the very tippy tip top of the curves in the other four states, we might just be seeing an inflection point. That is, the curve of new cases could be slowing from their near exponential rates of increase. The numbers released today we should expect to be lower than average. Consequently we will want to see the numbers beginning Tuesday through the end of the week to see whether this slowdown is real or a blip.
Regardless of whether or not new cases numbers are slowing down, we have to contend with rising numbers of deaths. Deaths of course lag new cases by weeks, sometimes as many as 4–8. So if we hypothetically hit peak new cases today, we would expect the number of deaths to continue rising and then peak perhaps sometime in mid- to late-January.
So where are we with deaths today? Also nowhere good. Let’s take a look.
Death curves for PA, NJ, DE, VA & IL.
In all five states with the potential exception of Illinois, new deaths continue to rise. Pennsylvania, worryingly, will likely surpass the peak death rate it saw in the spring if current trends continue. I would expect that sometime likely this week.
Illinois remains the one state where we might be seeing some good news. As I just mentioned above, deaths lag new cases by several weeks. And several weeks ago we appear to have peaked there in terms of new cases. It’s possible that we are beginning to or have already seen peak deaths in Illinois and that the next several weeks could be a gradual decline as the state gets its outbreak under control.
In the other four states, if we were to hypothetically peak with new cases this week, again, we would likely see these orange lines continue heading upwards for several weeks to come. And in that case, we’d almost certainly pass the peak death rates of the spring in Pennsylvania, Delaware, and Virginia. New Jersey might be the exception to that, however. And that would be largely due to the fact that so many deaths there happened so early in the pandemic before we had identified the best ways to save lives.
I suspect that the data coming out this week will be important to inform us whether or not we have crested or begun to crest this latest wave of infections.
Once more we look at the Covid-19 outbreak in Pennsylvania, New Jersey, Delaware, Virginia, and Illinois. And things are bad getting worse. I skipped last week because I was on holiday for Thanksgiving, but the data was perhaps not the most indicative of the current state of affairs at the time. But we now have a full week’s worth of data since the holiday, and like I said at the top, things are bad. Especially when we compare the charts below to where we were two weeks ago.
New cases curves for PA, NJ, DE, VA, & IL.
We look at new cases and we can see the impact of Thanksgiving on the shape of the curves. Note how in Pennsylvania, New Jersey, and to a lesser extent Delaware, we see a sharp plateau with the average before a sudden resumption of positive results. That’s Thanksgiving for you. You can see a similar, though perhaps more pronounced pattern in Illinois and Virginia where both states actually saw their average rate of new cases per day fall over the holiday.
Illinois, however, had been trending downward before Thanksgiving, and it might be due in part to the lockdown imposed by the city of Chicago. Whilst unpopular, lockdowns are an effective way of tamping down rising rates of new cases—vital to maintain capacity at hospitals and intensive care units (ICUs).
Of course cities and states are slowly implementing their own new lockdowns now. Philadelphia has been in one for two weeks now. Of course, we would want to wait 2–4 weeks to see if the numbers of new cases begin to fall, but the big intervening factor here is that very same Thanksgiving holiday. Did people travel? Anecdotally, I can say that the rooftop deck of my building’s parking garage, which I can see from my flat, was empty but for about five cars including my own. So people definitely travelled and likely visited other households. Not great.
And that could set us back, because new cases lead to new hospitalisations lead to new deaths. I would say that we probably should not expect as many deaths as we saw in the spring, because we are no longer dealing with a new virus. We know how to treat it far more effectively. But we also see that people aren’t taking the most basic preventative measures: wear a mask, and stay isolated.
In Illinois we are seeing death rates in excess of what we saw in the spring. And Pennsylvania isn’t far behind what we saw in March and April.
Death rates in PA, NJ, DE, VA, & IL.
And so I am increasingly worried that we will see more death in the winter than we did in the spring. And it’s depressing because so much of it could be avoided. Wearing a mask isn’t 100% effective. And there’s no guarantee that the few people an isolated household interact with, e.g. the delivery guy, aren’t themselves vectors. But both measures are far more effective than only occasionally wearing a mask at a house party to celebrate the holidays because we won’t let the virus beat us and interfere with our way of life.
The virus doesn’t care and you and I are tired of it. Tired of isolation. Of wearing masks. The virus is out there, spreading, and making people sick. And a fraction of those people are becoming ill enough to warrant hospitalisation. And a fraction of those are dying.
The next several weeks are going to be awful.
But you know that now. And you can brace yourself.
I have been taking and have yet to take a lot of holiday time this year. So apologies for the sporadic posting. But we’re working this week, because travelling to see family this year is a bad idea.
So the last Covid-19 update I posted was about a month ago. A lot has happened in the last month, like an entire election. But you really should go back a month and look at the charts for the five states I cover. At the time I said things were
Bad and getting worse.
I added that
while we are seeing dramatic rises in new cases, we are not yet seeing the rises in deaths that accompanied similar rises in March and April
And so let’s take a look at where we are now. First with cases.
New case curves for PA, NJ, DE, VA, & IL.
And now with those deaths.
Death curves for PA, NJ, DE, VA, & IL.
I mean…
This has all been so obvious for so long. And yet, I had to run two errands yesterday—timed so that I’d be running them whilst most of Philadelphia watched their football team and they must have played poorly from all the people yelling “c’mon” out their windows—and whilst the streets were fairly empty, about half of the people I passed either had their masks down, were doing that idiotic cover-the-mouth-but-not-the-nose thing, or—and this is the kicker—flat out had no mask on at all.
I’ll repeat what I said a month ago, things are bad and getting worse. But, maybe unlike a month ago, people will start taking this seriously. Because a month ago I wrote about how new deaths were not yet at the levels of the spring.
Well take a look at Illinois. They got there in just four weeks.
Pennsylvania? Halfway there.
New Jersey? Starting to rise a little bit faster now.
Virginia? Well Virginia has one of the odder death patterns I’ve seen—partly by their repeated cycling through backlogged data—but it’s clearly on the upswing now.
Delaware? It’s hard to see because the numbers are so small, but it’s also on the rise.
In case you weren’t aware, the US election is in less than a week, five days. I had written a long list of issues on the ballot, but it kept getting longer and longer so I cut it. Suffice it to say, Americans are voting on a lot of issues this year. But a US presidential election is not like many other countries’ elections in that we use the Electoral College.
For my non-American readers, the Electoral College, very briefly, was created by the country’s founding fathers (Washington, Jefferson, Adams, Franklin, et al.) to do two things. One, restrict selection of the American president to a class of individuals who theoretically had a broader/deeper understanding of the issues—but who also had vested interests in the outcome. The founders did not intend for the American people to elect the president. The second feature of the Electoral College was to prevent the largest states from dominating smaller states in elections. Why else would Delaware and Rhode Island surrender their sovereignty to join the new United States if Virginia, Pennsylvania, and New York make all the decisions? (The founders went a step further and added the infamous 3/5 clause, but that’s another post.)
So Americans don’t elect the president directly and larger states like California, New York, and Texas, have slightly less impact than smaller states like Wyoming, Vermont, and Delaware. Each state is allotted a number of Electoral College votes and the key is to reach 270. (Maybe another time I’ll get into the details of what happens in a 269–269 tie.) Many Americans are probably familiar with sites like 270 To Win, where you can determine the outcome of the election by saying who won each state. But, even though the US election is really 50 different state elections, common threads and themes run through all those states and if one candidate or another wins one state, it makes winning or losing other states more or less likely. FiveThirtyEight released a piece that attempts to link those probabilities and help reveal how decisions voters in one state make may reflect on how other voters decide.
The interface is fairly straightforward—I’m looking at this on a desktop, though it does work on mobile—with a bunch of choices at the top and a choropleth map below. There we have a continually divergent gradient, meaning the states aren’t grouped into like bins but we have incredibly subtle differences between similar states. (I should also point out that Maine and Nebraska are the two exceptions to my above description of the Electoral College. They divide their votes by congressional district, whoever wins the district gets that Electoral College vote and then the state overall winner receives the remaining two votes.)
Below that we have a bar chart, showing each state, its more/less likely winner state and the 270 threshold. Below that, we have what I’ve read/heard described as a ball plot. It represents runs of the simulation. As of Thursday morning, the current FiveThirtyEight model says Trump has an 11 in 100 chance of winning, Biden, conversely, an 89-in-100 chance.
But what happens when we start determining the winners of states?
Well, for my non-American readers, this election will feature a large number of voters casting their ballots early. (I voted early by mail, and dropped my ballot off at the county election office.) That’s not normal. And I cannot emphasise this next point enough. We may not know who wins the election Tuesday night or by the time Americans wake up on Wednesday. (Assuming they’re not like me and up until Alaska and Hawaii close their polls. Pro-tip, there’s a potentially competitive Senate race in Alaska, though it’s definitely leaning Republican.)
But, some states vote early and/or by mail every year and have built the infrastructure to count those votes, or the vast majority of them, on or even before Election Day. Three battleground states are in that group: Arizona, Florida, and North Carolina. We could well know the result in those states by midnight on Election Day—though Florida is probably going to Florida.
So what happens with this FiveThirtyEight model if we determine the winners of those three states? All three voted for Trump in 2016, so let’s say he wins them again next week.
We see that the states we’ve decided are now outlined in black. The remainder of the states have seen their colours change as their odds reflect the set electoral choice of our three states. We also now have a rest button that appears only once we’ve modified the map. I’m also thinking that I like FiveyFox, the site’s new mascot? He provides a succinct, plain language summary of what the user is looking at. At the bottom we see what the model projects if Arizona, Florida, and North Caroline vote for Trump. And in that scenario, Trump wins in 58 out of 100 elections, Biden in only 41. Still, it’s a fairly competitive election.
So what happens if by midnight we have results from those three states that Biden has managed to flip them? And as of Thursday morning, he’s leading very narrowly in the opinion polls.
Well, the interface hasn’t really changed. Though I should add below this screenshot there is a button to copy the link to this outcome to your clipboard if, like me, you want to share it with the world or my readers.
As to the results, if Biden wins those three states, Trump has less than a 1-in-100 chance of winning and Biden a greater than 99-in-100.
This is a really strong piece from FiveThirtyEight and it does a great job to show how states are subtly linked in terms of their likelihood to vote one way or the other.
Credit for the piece goes to Ryan Best, Jay Boice, Aaron Bycoffe and Nate Silver.
Yep, Covid-19 remains a thing. About a month or so ago, an article in City Lab (now owned by Bloomburg), looked at the data to see if there was any truth in the notion that people are fleeing urban areas. Spoiler: they’re not, except in a few places. The entire article is well worth a read, as it looks at what is actually happening in migration and why some cities like New York and San Francisco are outliers.
But I want to look at some of the graphics going on inside the article, because those are what struck me more than the content itself. Let’s start with this map titled “Change in Moves”, which examines “the percentage drop in moves between March 11 and June 30 compared to last year”.
Conventionally, what would we expect from this kind of choropleth map. We have a sequential stepped gradient headed in one direction, from dark to light. Presumably we are looking at one metric, change in movement, in one direction, the drop or negative.
But look at that legend. Note the presence of the positive 4—there is an entire positive range within this stepped gradient. Conventionally we would expect to see some kind of red equals drop, blue equals gain split at the zero point. Others might create a grey bin to cover a negative one to positive one slight-to-no change set of states. Here, though, we don’t have that. Nor do we even get a natural split, instead the dark bin goes to a slightly less dark bin at positive four, so everything less than four through -16 is in the darker bin.
Look at the language, too, because that’s where it becomes potentially more confusing. If the choropleth largely focuses on the “percentage drop” and has negative numbers, a negative of a negative would be…a positive. A -25% drop in Texas could easily be mistaken with its use of double negatives. Compare Texas to Nebraska, which had a 2% drop. Does that mean Nebraska actually declined by 2%, or does it mean it rose by 2%?
A clean up in the data definition to, say, “Percentage change in moves from…” could clear up a lot of this ambiguity. Changing the colour scheme from a single gradient to a divergent one, with a split around zero (perhaps with a bin for little-to-no change), would make it clearer which states were in the positive and which were in the negative.
The article continues with another peculiar choice in its bar charts when it explores the data on specific cities.
Here we see the destinations of people moving out of San Francisco, using, as a note explains, requests for quotes as a proxy for the numbers of actual moves. What interests me here is the minimalist take on the bar charts. Note the absence of an axis, which leaves the bars almost groundless for comparison, except that the designer attached data labels to the ends of the bars.
Normally data labels are redundant. The point of a visualisation is to visualise the comparison of data sets. If hyper precise differences to the decimal point are required, tables often are a better choice. But here, there are no axis labels to inform the user as to what the length of a bar means.
It’s a peculiar design decision. If we think of labelling as data ink, is this a more efficient use with data labels than just axis labels? I would venture to say no. You would probably have five axis labels (0–4) and then a line to connect them. That’s probably less ink/pixels than the data labels here. I prefer axis lines to help guide the user from labels up (in this case) through the bars. Maybe the axis lines make for more data ink than the labels? It’s hard to say.
Regardless, this is a peculiar decision. Though, I should note it’s eminently more defensible than the choropleth map, which needs a rethink in both design and language.
