Census data fascinates me from a data visualisation perspective; one can look at it so many different ways. Last week I looked at some of the Slovakian census data on the Carpatho-Rusyns that live in the northeastern mountains of Slovakia. But yesterday, the British Office of National Statistics released the results from their census of England and Wales (Scotland reports later and Northern Ireland did so already, yay devolution.) One of the big news stories was that England and Wales had 500,000 more people than had been expected. That doesn’t sound like a lot of people, but to put it roughly into American proportions, that would be like finding that there was a whole new city the size of Chicago somewhere in the United States.
But while many organisations and individuals will certainly be looking at the census data in the coming days, weeks, and months, the ONS released its own interactive application. Basically it looks at the population pyramid for England and Wales from 1911 to 2011, a century’s worth of data. But what makes this different from the GE population pyramids, for example, is the context that the ONS has added that strict data pulls lack.
Here in 1921, rolling over a particular cohort reveals the details of those aged 30 in 1921. There is a clear difference between the number of men and women. But why? The text block’s first note details how 700,000 men aged 20–40 died during World War I and thus altered the basic structure of the English and Welsh population.
And in 1951 we begin to look at the British baby boom in the post-war era. Again, while the Baby Boom might be expected, the ONS also points out that the NHS, the British National Health Service, had also recently started and was positively affecting life expectancy and the general health of the British public. These are again things that would not likely appear in more data-focused pieces.
But everybody loves to compare things to other things. So, the ONS also released a more data-focused application that allows the user to select two different census geographies and compare them. This is more as one would expect, comparing overlays vs. side-by-side looks at different population pyramids. The example below compares London to Birmingham.
Credit for the pieces go to the ONS Visualisation Centre.
As the Supreme Court is likely to scrap the mandate provision of the health care law—without which sick people are left to pay higher premiums if they can get coverage at all—later today, the New York Times looks at the impact of removing the health care law changes the number of people without health insurance.
It appears as if the Greeks, who voted in parliamentary elections for the second time in as many months, have narrowly voted for pro-bailout parties. But whether the pro-bailout parties can put aside their other political differences and form a coalition government remains to be seen.
I appreciate the mirror approach, but wonder if the comparisons might not have been clearer if measured directly? Or what would have happened without the mirror approach and compared the two countries in single but slightly larger charts? Regardless, one can easily see that Greece has some serious problems.
Credit for the piece goes to Andrew Barr, Mike Faille, and Richard Johnson.
This falls under the just-because-it’s-about-geographies-doesn’t-mean-it-should-necessarily-be-visualised-as-a-map category. The Guardian has taken data from the African Economic Outlook, specifically real GDP growth rates, and charted them as a map. This caught my interest initially because of some work I have been doing that required me to read a report on African economic development in coming years. So I figured this could be interesting.
But it’s a map. That’s not to say there is anything inherently wrong about the map. Though the arrangement of the legend and size of each ‘bin’ of percentage values is a bit odd. I would have placed the positive at the top of the list and tried to provide an equal distribution of the data, e.g. 3–10 for both positive and negative values. But, without looking in any depth at the data, the designer may have had valid reasons for such a distribution.
That said, two finer points stick out to me. The first is Western Sahara. Long story short, it is a disputed territory claimed by different factions. I am not accustomed to ever seeing any real economic data coming out of there. But, according to the map, its growth is 0–3%. When one looks at the data, however, one finds that as I would have expected the data says “no data”. Ergo the green colour on the map is misleading. Not necessarily incorrect, for the growth could have been between those two points, but without any data one cannot say for sure.
The second concern for me is South Sudan—remember that story? For starters one cannot find it on the map; South Sudanese territory is depicted as part of Sudan. While South Sudan is one of the poorest countries on the earth, its split from Sudan is rather important. Looking at the data, one can see Sudan’s growth went from 8 to 4.5 to 5 to 2.8. Why the sudden drop? Probably because Sudan’s economic boom has largely been built on the boom in oil prices over the past decade or so. But, most of that oil is no longer in Sudan, Not because its been pumped dry, but rather most of the oil fields can now be found in South Sudan.
These are some of the contextual stories that make sense of a data set. But these are the stories lost in a simple, interactive map.
The Globe and Mail has been working on a story about immigration to Canada because apparently not all immigrants come to America. The story has its section headers running down the side column of the page, like many other segmented stories you’ll see posted online these days, but also uses graphics to make and supplement its arguments.
This one chart from the piece is an example of how the simple format of a line chart can clearly express and visualise an interesting trend. Immigrants from the past two decades earn less than immigrants to Canada in the 1970s. Those from the early 90s, however, do appear to have a faster rate of income growth that approaches parity with Canadian-born income-earners.
On Sunday the New York Times featured a small graphic highlighting the disparity in growth rates across the G-20 if broken into the ‘core’ G-8 and then what one might call the emerging markets of the G-11.
The charts are small yet compelling in telling the story of how the two different groups are performing. However, I was left wanting to better understand the comparisons between the sizes and growth of the various countries. The areas of circles are difficult to compare and aggregates mask interesting outliers. So, using what I imagine to be the same data from the IMF, I took a quick try at the data to create my own infographic.
Indeed, interesting stories began to appear as I plotted the data. Russia is a member of the G-8, but perhaps has more in common with the G-11. After all, Russia’s growth was nearly 500%. Similarly interesting were Canada and Australia. The former, a G-8 country, was the only G-8 country besides Russia to have greater than 100% growth. And Australia, certainly not an emerging market in most senses, experienced nearly 300% growth. Whereas the emerging markets of Mexico and South Korea lag behind the rest of the G-11.
Then, when plotting the sizes of the economies, China was no surprise as the second-largest economy. However, that Brazil has managed to already surpass the G-8 economies of Italy, Russia, and Canada was a bit shocking. And Brazil looks nearly ready to surpass the UK, but for its apparent recent downturn. Also interesting to note are the Financial Crisis dips in GDP across most countries. Some countries, like China, unsurprisingly did not suffer greatly. However, that Japan and South Africa kept on a steady pace of growth was unexpected.
All of that would have been missed but for a slightly deeper dive into the IMF data. And a few hours of my time.
This weekend the New York Times looked at segregation in New York City schools by mapping the least (and most) diverse and offering quick comparisons to other large cities. (Is it really a surprise that the country’s largest cities also would need the largest demographic shifts to create diverse education environments?) Probably the best thing, seemingly as always, in the piece is the annotations that provide stories and context and explain the outliers that are all otherwise visualised in the infographic.
Mariano Rivera of the New York Yankees is(was?) one of the best closers in baseball history. I’ll give him that. So when a freakish accident brought to an end his season—and possibly his career—the New York Times of course had an(other) infographic about his historic numbers.