Visualizing Unemployment

So, as I may have mentioned before, I am currently taking a MOOC on infographic and data visualization with the Knight Center for Journalism in the Americas, at the University of Miami, taught by the man himself, Alberto Cairo. I will have a full reflection post on the experience once the course is over. For now, I’ll just say that taking this course was the best idea I had this term.

Anyhoo… since I am spending hours on work for this course, I thought I’d share it here, but you guys aren’t allowed to make fun. Heck, I pulled an all-nighter last night, it felt like being back in the good/bad old days of dissertation work!

[Click on all the images for ginormous views]

So, this week’s exercise was based on a post by the Guardian’s Datablog (here) and critically examine the data visualization, then, let our imaginations run wild (ok, maybe not) and come up with some alternative or expansion or whatever on the topic of visualizing unemployment. So, off to downloading data in Calc I went (I use OpenOffice because I’m cool) from the Bureau of Labor Statistics to start playing with them. Here is the static version of the Guardian’s interactive map:

From there, I decided that it might be nice to have slightly different callout boxes from each state, different than the one from the Guardian. Especially, I thought state time series would be nice. My second idea, for the US, was to get some data to compare the official unemployment rate (AKA: U3), used above and commonly, to the total unemployment rate (AKA: U6, less well-known). I got the data, went into Tableau and produced this interactive bar chart (embedding does not work).

Here is the static version:

Pretty striking differences.

I then went back to Tableau to get a time series on the contrast between U3 and U6. And here is the static version:

Ok, so, you see where this is going. There are political points to be made here and they are pretty obvious.

So, when I put it all together, this was part 1 of my assignment:

For this one, above, I used what is still one of my favorite software: Simple Diagrams, for the canvas and the placing of the different items. You see the general US map, then a few sample callout boxes by states (and you can see that the unemployment lines look very different), and then my U3 v. U6 comparisons.

Then, for part 2, I decided to go for some international comparisons. I downloaded some more data, back to Calc, back to Tableau and the result was this line graph comparing countries (still no embedding, sorry). The static version looked like this:

From there, I derived three patterns and one country in a class of its own (Germany) and produced the relevant view. Here is one example:

So, I took all my 3 patterns + 1 and put them in an infographic using Piktochart (you really need to click on this one for the truly ginormous view):

And that’s it, folks.

Obviously, I spend so much time on the data processing / data visualization part that I don’t really think about the “story”. Once I’ll have a greater mastery of all the software stuff, it will get easier (and quicker! I’m too old for all-nighters!).

See, the issue is that I was taught statistics before all the fancy software. And even when visualization software came along, it did not seem to matter because graphics were part of research papers, destined to peer-reviewed publications where aesthetics really does not matter. I mean, look up any journal and take a look at the visualizations (if there are any). It’s dry, drab, grey and sad. So, I never learned this stuff. This means that I have to spend way more time than some of my classmates on the visual part and the content part kinda takes a back seat because that is less my priority.

5 thoughts on “Visualizing Unemployment

  1. Now the interesting questions that come from this data, what is Germany doing that is working so well that they others are not? Another interesting question is what is it about Japan that kept their unemployment so low and relatively consistent?

    • Each pattern deserves an analysis of its own. Of course, there would be multiple factors: regulation levels, or deregulation (especially in the financial sector), levels of social spendings, labor protections, etc.

  2. I can relate to the introduction of this post so much!! being your colleague in the course 🙂 I like the angle of comparing countries besides the US too. It could go somewhere 🙂

    Great job 🙂 you have used gamut of softwares I should explore too. Please do tell me which forum you submitted your work in, so I could comment there too


  3. Pingback: The Global Sociology Blog - The MOOC Experience

Leave a Reply

Your email address will not be published. Required fields are marked *