View the project live at http://www.superdiv.mmg.mpg.de/
Eric Rodenbeck: We’re here with Stamen’s Logan Williams talking about our recent work for the Max Planck institute visualizing super-diversity. What’s the project about?
Logan Williams: This was a project initiated by professors at the Max Planck Institute and the University of British Columbia, to show how the nature of diversity in our urban areas around the world has been changing over the past few decades.
We’re now collecting more data and have a better understanding about how and why people immigrate. People coming into a country aren’t all coming from the same place. Even those coming from the same place aren’t all coming for the same reason or with the same background.
ER: Noah Veltman, who made a beautiful map of the history behind the place names of San Francisco streets, told me an interesting story about Chinatown in New York. The Chinese names for the streets, which are different than their English names, are reflective of the different parts of China where those people come from. From the outside, it can seem like Chinatown is this homogenous group of Chinese people. But in fact, within their own culture, they’re from all over the county and highly varied, and they separate themselves out according to where they’re from.
LW: Right. What I don’t understand yet is how much of the nature of super-diversity is due to the nature of migration changing, and how much of it is due to the nature of data collection and data bureaus and statistics changing. That’s a question for the professors.
ER: If you look at the American Panorama Project, you can see the data get more refined over time in the census. Even as recently as in 1970, your census taker would have just said, “You’re from Africa.” But in 1990, they started to care about whether you were from Egypt or Nigeria. So that’s sort of a new part of it, and you can see it in the data. But there’s a premise here that the kind of diversity that we’re seeing now irrespective of the statistical differences is of a different type than has been before. Is that part of the premise?
LW: Yes. Diversity itself is becoming more diverse in a lot of ways.
ER: And these cities are very particular in that sense. They’re not cities that are necessarily seen as titans on the world stage like Tokyo or New York. They’re real cities, though: Vancouver, Sydney. They were chosen for this study because they’re global cities in a particular way.
So we’re taking a look at the first graph, and these are the changing sources of permanent migration from three parts of the former British Empire — Canada, Australia, and New Zealand — and their percentages. So you can start to see much larger numbers into New Zealand from the Americas, where previously it was dominated by Oceania.
LW: This initial visualization was one of the simplest for this project. It leads because it immediately shows the depth of the data that’s behind the project over a multidecade timespan, and because I think it raises some interesting questions. Just looking at this data, you might wonder, Why was there such a jump in European migration in the early ’80s in Canada? Or what exactly changed in New Zealand’s immigration policy in the late ’90s to create this dramatic effect?
ER: Right. So it’s about the policy of the country, but then potentially also about things that are happening in the world as well that would cause people to move. Cool.
Then we move onto the national scale. And here we’re looking at Canada immigration by category from 1980 to 2017. So what are we showing here?
LW: This visualization shows the makeup of origin of immigrants to Canada or to the other two cities that we’re looking at based on the type of visa that the migrant arrived with. To me, the interesting thing about this one is the ability to see these specific global events that you were talking about, for instance, if you look at the humanitarian immigrants in the early ’90s there.
So as you click, if you progress year by year, you can see different refugee crises up here and recede from the perspective of Canada.
ER: Sri Lanka seems to have a large number of people pretty consistently.
LW: And that would be the ongoing civil war, for example. But at the other periods of time — for instance, in the mid-’90s you can see the Bosnian War. Or certainly, in the 2010s, you can find the Syrian civil war pretty easily.
ER: I can see that between 2014 and 2015 the number of Syrian immigrants triples within a single year. And this chart is showing the in-migration, right, and not the number of people living in those places?
LW: Right. This is the in-migration in that particular year.
ER: And so presumably you can tell the same kind of thing in the economic immigration, right?
LW: Right. Especially if you look over a long time scale, the nature of people coming for economic reasons to Canada changes with time.
ER: You can see the number of people from Hong Kong, I’m just watching it go down as I move from the 1990s to 2000s.
LW: Maybe that’s a policy change in Canada. Maybe that’s also reflective of increasing economic opportunities in Hong Kong. I don’t know.
ER: Right. And again, I think the visualizations are not necessarily designed to give you the answers, but to at least give you some specifics to ask a question about.
LW: Right. And specifically, they’re designed to support the actual experts in this.
ER: So now we’re looking at New Zealand.
LW: You can see that until the mid-2000s, there wasn’t a differentiation between these different categories and the data that we have.
ER: And interestingly, I don’t see a category for humanitarian data.
LW: New Zealand’s kind of the odd one out in all of these because they used a pretty different method of categorizing people because origin and nationality are very complex. There are multiple ways to think about it.
ER: And there’s more sheep than people in New Zealand, I think. So this is basically a tale of the relationship between migration time and world events. Cool. So then we get to the metropolitan one. It’s the most interesting one, I think, of the visualizations that we’ve come up with.
LW: This was one of the most challenging and most fun visualizations to make. The goal here was to show how the connection between what religion people practice and what country they’ve migrated from has changed with time.
ER: Okay. So we can see, for example, that the Australian population of Sydney in 2016 is about evenly distributed between Western Catholicism, the Anglican Church of Australia, and no religion at all, plus a smattering of other little religions.
LW: Right. And then one of the things that was fun to design with this visualization was the colors. If you look at those colors, you can actually see a kind of implicit bar chart that’s made next to Australia showing that it is about two-thirds Christian and about one-third no religion.
This was intended to help the researchers at the Max Planck Institute explore this idea that not only have the countries that people have been coming from to these three cities changed over time, but the makeup of the people who are migrating has changed. That the religion practiced by Ukrainian immigrants to Canada might have gotten more diverse over time, as well.
ER: They’re about evenly split in 2011 between no religion and Roman Catholic, whereas in ’91, a lot more Roman Catholics and Christian Protestants. So the population of Ukrainian immigrants has gotten less Protestant over time.
LW: I can’t speak for why that happened, necessarily, or what that means for government or sociological policy. But it was certainly an interesting visualization to create. And there are so many threads in this that I think it’s one of the deepest visualizations in this piece to keep exploring.
ER: Let’s talk about the color for a moment, because it was a surprise to me that you and Alec inverted the color gradients the way that you did. I think it’s great.
LW: We used a set of colors for countries and a set of colors for religion. And then the connection between religion and country was a color gradient. But we actually discovered pretty early on that if we reversed that color gradient, it let us create all of these kind of implicit bar charts of where people were coming from for a particular religion, or what religion people were practicing for a particular country or region.
ER: I love that. I’m looking at Australia now. You don’t even have to look over to the side to know that it’s a lot of Western Catholics and Anglican Church of Australia. But then the relationship goes back the other way, too. So a lot of really dense information in there.
ER: Looking at the intersections in this visualization, thinking about the relationship between diversity and opportunity, the use of sentences in data visualization is one of my favorite things. You’re adjusting a sentence using these dropdowns, which gives it a narrative component and tells you some really basic things. One of the things I’ve found in data visualizations over the years is that there’s no reason to not use text to describe what people are looking at and give them a sense of what’s there. It doesn’t have to just all be from the visuals.
LW: We’ve done a lot of looking at broad scale data patterns. This was finally something that was much more personal, which we conveyed by thinking about it as a sentence and using the first person in these sentences. I think it helps connect the dots between statistical patterns and actual people who are migrating, and trying to find economic success and a life in a new place.
ER: You can see that for people who immigrated before 1980, and for cases where their grandparents immigrated to Australia when they were 24 years old, there isn’t any impact on their housing. But presumably, when you’re older, that would change.
LW: We can start to do things like look at the relationship between immigrants who immigrated before 1980, and those whose parents migrated, and those whose grandparents immigrated. I think the conclusions are that as people stay in a country longer, they start to look much more like the general population than they do their parents and grandparents.
ER: I don’t want to say that I know what conclusions an academic with domain expertise might draw from this. But what I find interesting about this is the intersectionality of everything. That your class, economic, and social status in a country isn’t a factor of a single variable, but a factor of multiple variables all intersecting and combining together. Data visualization is a great way to show that.