These are some of the data stories that caught my attention this week:
This nytimes data viz could be overwhelming because they include so many different kinds of contraception. But because they’re jiggered it so that, as you change your position on one graph, it automatically changes your position on all the other graphs. This makes it really easy and fun to compare the chance of getting pregnant by each method (which they could do with a single bar chart) but also show how these rates are magnified over time, and how perfect use compares to normal use. Well done.
This nytimes data viz breaks down county-level election results, by both voter-density and party preference and is notable for the elegance of its design. Although it’s not particularly innovative, the soft gradation changes allow you to really see how certain cities and states break down. For instance, although New Orleans is very Democratic, the other cities in Louisiana are notably split, which goes contrary to the typical urban/rural split. Whereas in Arkanas–Little Rock is almost exclusively Democratic and the rest of the state is mostly Republican–although you can also see the differences between Fayetville, a University town in Northwest Arkansas, and its neighbor Rogers, the home of Wal-Mart.
This was my favorite data story of the week because Nate Silver breaks down the very unusual election result in a Virginia Senators successful election race–a candidate who looked to have a 10-point lead by most polls, but ended up barely squeaking by. After breaking the issue down and explaining how vicissitudes in Virginia’s voter turnout can help explain the anomaly–he explains that really, it’s so rare, that it’s more comparable to a couple of races from twenty or thirty years ago. And then he provides a theory to explain it all–that in a rare race like this, (as in Bloomberg’s 2009 election) voters sometimes cast a ‘protest’ vote, where they want the candidate to win but they also went to send a message. It would be interesting to figure out some way to measure this, because you gotta figure there were some protest votes against Cuomo in this most recent election, even though few really wanted Astorino to win. Silver has been so good about explaining polling data–he had another column on how the polls skewed democratic this year, and compared it to year’s past, which was also very good.
The story I want to focus on, though, is more of an enterprise story, without a real news peg. It shows how a couple’s life satisfaction changes after one child and two children. But crucially it compares how men and women differ between the first and second child and the results are startling. The first chart very quickly shows that more women’s happiness goes down after the second child than the first. The second chart shows that men’s relationship satisfaction goes down substantially for baby 1 and 2, but strangely women’s relationship satisfaction goes up. The article doesn’t offer a good explanation here, of how it’s possible that women can be both less happy overall with two children, but more happy with their relationship. Are they happy that they are in a family now with now children, even if it’s exhausting and tiring?
Screen Shot 2014-11-10 at 11.54.01 PM
The third chart shows an even more perplexing trend. Having a second child doesn’t negatively affect a couple’s views about their financial situation. Part of this could be that the biggest change came after the first child and, as the article points out, a second child costs relatively less. But what’s perplexing is that the extra financial burden bothers men less than it does women. Is this because, as women, they’re losing financial power in the relationship, by taking off work, which causes them to be more dependent on their husband and fall behind in their career advancement? Could it be that this explains the fall in women’s happiness in the first chart–that, though they’re relationships are better, its offset by the losses they experience professionally?
The writing in this post doesn’t add much to the vizualiation–it basically just re-hashes the numbers, explains how they were derived a bit and gives some personal insight. We don’t need as much explanation of how she got the numbers in the main text. It would’ve been better to hear some expert explanations try to make sense of the data.
It also could use a little more cohort analysis. The most gaping questions are about how much overlap there is between the people who changed their minds between the first and second baby. Is there a large cohort whose satisfaction never changes, no matter what happens with the children? Or are the people who are unhappy about life also the one’s that are unhappy about their financial situation? Or is it the relationship changes that are affecting overall happiness? If the proportion of men who are less happy in their relationships and less happy with their finances is greater than the number of men who are less happy overall–why isn’t the decrease in relationship and financial satisfaction impacting the overall happiness of some men? Or might there be quirky clusters–women for whom, relationship satisfaction affects overall happiness, but a drop in financial satisfaction has little effect–and others for whom the exact opposite is true?
But overall I really liked this visualization because its results are so quickly discernible, contrary to received wisdom and provoke a lot of interesting follow-up questions.