But at least today is Friday. Also Cinco de Mayo. And so in addition to trying to find some mezcal tomorrow—anybody know a good spot in Philly?—we can wrap this week up with something on the humourous side.
My mobile is a few years old now and I’ve been taking lots of the photos the last few years. Last weekend I reached a point where I could no longer take photos. Consequently I have been going back through all my old photos. And so this piece from xkcd seemed rather appropriate.
Well as of last night, we are having yet another vote on AHCA, better known as Trumpcare. I will not get into the details of the changes, but basically it can be summed up as waivers for Obamacare regulations. And as of last night, $8 billion over five years to cover those at high-risk. What about after five years? What if, as experts say, that sum is insufficient and it runs out before five years are up?
This is still a bad bill.
But thankfully we have FiveThirtyEight who looked at support before the Upton amendment—the $8 billion bit—and found that the bill could still fail because of a lack of moderate support.
The basic premise is this: In order to get the conservative Freedom Caucus, which scuppered the bill a few weeks ago, on side Ryan et al. had to make the bill more conservative. They likely had to make it cover fewer people at a higher cost. I say likely because Ryan is not sending this to the Congressional Budget Office (CBO) to score the bill, something typically done to see how much it costs and whether it might work. Problem is, by making the bill more conservative, they push away moderate Republicans. Yes, Virginia, they do exist.
Today’s question is whether an $8 billion throw-in will buy in enough moderate votes.
Another day, another story about the administration to cover with data-driven graphics. We are approaching Trump’s 100th day in office, traditionally the first point at which we examine the impact of the new president. And well, beyond appointing a Supreme Court justice, it is hard to find a lot of things President Trump has actually done. But on his 99th day, he will also need to approve a Congressional bill to fund the government, or else the government shuts down on his 100th day. Not exactly the look of a successful head of state and government.
Why do I bring this up? Well, one of the many things that may or may not make it into the bill is funding for Trump’s wall that Mexico will pay for, but at an undetermined later date, because he wants to get started building the wall early, but late because he promised to start on Day 1.
Several weeks ago the Wall Street Journal published a fantastic piece on the current wall bordering Mexico. It examines the current state of fencing and whether parts of the border are fenced or not. It turns out a large portion is not. But, the piece goes on to explain just why large sections are not.
You should read the full piece for a better understanding. Because while the president says building the wall will cost $10 billion or less, real estimates place the costs at double that. Plus there would be lawsuits because, spoiler: significant sections of the border wall would cross private property, national parks, and Native American reservations. Also the southern border crosses varied terrain from rives to deserts to mountains some lengths of which are really difficult to build walls upon.
But the part that I really like about the piece is this scatter plot that examines the portion of the border fenced vs. the number of apprehensions. It does a brilliant job of highlighting the section of the border that would benefit most significantly from fencing, i.e. a sector with minimal fencing and a high number of apprehensions: the Rio Grande Valley.
And to make that point clear, the designers did a great job of annotating the plot to help the reader understand the plot’s meaning. As some of my readers will recall, I am not a huge fan of bubble plots. But here there is some value. The biggest bubbles are all in the lower portion of fenced sectors. Consequently, one can see that those rather well-fenced sectors would see diminished returns by completing the wall. A more economical approach would be to target a sector that has low mileage of fencing, but also a high number of apprehensions—a big circle in the lower right of the chart. And that Rio Grande Valley sector sits right there.
Overall, a fantastic piece by the Wall Street Journal.
Credit for the piece goes to Stephanie Stamm, Renée Rigdon, and Dudley Althaus.
We have a scatterplot from the Financial Times that looks at wage and economic growth across the OECD, focusing on the exception that is the United Kingdom. And that is not an exception in the good sense.
The UK had the rare privilege of experiencing economic growth—that’s good—while simultaneously wages fell—that’s bad. But I wanted to comment on the chart today.
Straight off the bat, the salmon-coloured background does not bother me. That is FT’s brand and best to stick to it and make your graphics work around it. Possibly the colours in the plot could use a bit of a push to increase separation, but that is more a design quibble. Instead, I am not too keen on the colour coding here.
Not that the colours need not be applied, but why to the dots? Note how the dots of a colour fall into one of the quadrants. Instead of having people refer to the legend, incorporate the legend into the chart by moving the labels to the plot background. You could colour code the labelling or even colour the quadrants to make it a bit clearer.
Credit for the piece goes to the Financial Times graphics department.
So this isn’t quite a shocker, but the BBC gained access to more granular Brexit vote data, and then examined the results against demographic data. The conclusion, a lower education level best corresponded to voting to leave the European Union. Again, we all sort of knew that, but this provides an even larger, richer sample size.
What is interesting from the American perspective is how that compares with the election of Donald Trump. In that case as well, lower levels of education correlated well with votes for Trump.
Of course now I will be closely following the elections in the Netherlands, France, and Germany this year to see if the same lower education level corresponds to the vote in favour of populist, nationalist political parties, e.g. Marine le Pen and Geert Wilders.
Credit for the piece goes to the BBC graphics department.
Today’s post is a choropleth map from the Washington Post examining diversity in the United States and how fast or slow diversity is expanding. Normally with two variables one goes instantly to the scatter plot. But here the Post explored the two variables geographically. And it holds up.
The colours are perhaps the only part holding me up on the piece’s design. Are blue and yellow the best two colours to represent level of diversity and growth? I lose some of the gradation in the yellows, especially between the big increases in diversity. Can I offer a better solution? No, and maybe there is not. But I would love the chance to explore different palette options.
As you well know, I am not a big fan of always plotting things on maps. I call them the silver bullet. However, in this instance, there are clear geographic patterns to the four different scenarios. Of course this soon after the election I would love adding a third variable: how the counties voted in the presidential election. Maybe next time.
Credit for the piece goes to Dan Keating and Laris Karklis.
Happy Friday after the election. Now that we have had our fill on sweets and bitters, we probably need to move towards a more balanced, more moderate diet. A couple of months ago the New York Times put together this scatter plot from the difference between public and nutritionist opinion on whether certain common foods are healthy.
I normally do not comment on the design of my Friday posts, since I intend them to be on the lighter, more humourous side of things. But this piece interests me, because despite the seriousness of the subject matter I find it lighter and less serious. Why? After studying it, I think it is because of the inclusion of photographs of the items. With the labels still present, I am left thinking that a small dot would be equally effective in communicating what falls where.
But more importantly, look at the sizes of the images relative to the plot. Take the bowls of granola or popcorn, for example. They occupy almost an entire square; the actual value could be anywhere with the 10 percentage point range either vertically or horizontally. And for those two, it does not matter a great deal. Each falls firmly on one side of the line. But what about butter? Kind bars? Cheddar cheese? The large graphic size straddles the line, but because the designers opted for photos over more precise dots, we cannot ascertain whether these foods fall on one side of the line or the other.
The point is that the graphics and design of a piece can influence the perceived seriousness of a piece. An image of a can of Coca-Cola certainly can be more engaging than a 10-pixel dot. But the precision of the dot over the image can also be engaging to the right audience, an audience interested in the data behind the story. There are ways of integrating both, because later on in the same article, we see a means of doing just that.
Here the image provides supplemental information. Just what does a granola bar look like? Well here you can see it. But even here, despite the smaller size and cropped dimensions, the photographs steal a bit of emphasis from the numbers and the charts to the right. (For things like SlimFast, that is no surprise, because the package is designed to capture your attention.)
At the end of the day, the piece interests me because the data interests me. And the story interests me. And I generally like the data visualisation forms the designers chose. But I keep getting hung up the photographs. And not in a good way. What do you think? Do the photos add to the story? Do they make the data clearer?
Credit for the piece goes to Kevin Quealy and Margot Sanger-Katz.
Among my recent life changes, I have moved back to Philadelphia from Chicago. That means Sundays the bars and streets are full of people wearing the jerseys of their favourite Eagles (American) football players. And apparently this year, they are off to a good start. FiveThirtyEight took a look at other 3-0 teams to compare the 2016 Eagles and examine their chances for the playoffs.
Today’s post is about religion. One of the two things you are never supposed to talk about in good company. And since the other is politics and since I cover that here frequently, let’s just go all in, shall we?
FiveThirtyEight has an interesting piece about religious diversity and a corresponding lack of religiousness. From a graphics standpoint, the central piece is this chart below.
What I would love, however, is for the plot to be interactive. It would be great to let people check out their own individual home states and see how they compare to the everyone else.
Credit for the piece goes to the FiveThirtyEight graphics department.
Last week we witnessed the lorry attack in Nice, France. This week we have the axeman attack on a German train. Does anybody note, however, the recent terror attacks in Dhaka, Bangladesh? Probably not, according to this insightful piece from FiveThirtyEight. They took a look at journalism’s coverage of terror attacks and whether there are discrepancies based on geography. Turns out that yes, there are. But, the article does make a point to note some reasons why that might be. One, we have covered it a lot more often since 11 September 2001. Anyway, the whole piece is worth a read.
Credit for the piece goes to the FiveThirtyEight graphics department.