Anything other than the bottom value is hard to figure out what the trend is over time, showing they add up to 100% is kind of a waste. The x axis ticks are not equidistant (10, 5, and 2 years between columns)..
Thanks for posting the interesting article. Wish the graph was better.
It may be awful from a professional point of view, but it's one of the few popular media graphs where I (a non-professional) can immediately see interesting changes.
I think the decreased period of the x on the right is useful, because we're probably more interested in details in the near past than in 1970. And I don't think it's necessary to expect high fidelity in a popular article meant to summarize a lot of data for lay people.
The numbers, and the relative size of blocks, mitigate your point about trends being difficult to see. For example, it's easy for me to see that both types of marrieds were about 70% in 1970, and about 50% in 2012.
It's also easy to see that, in percent:
- Marrieds with children show a major decline.
- Marrieds without children are steady (and I really wonder why that would be, if it's not coincidental).
- Total marrieds show a major decline.
- Households with children show major decline (married with plus other families).
- All other categories, which are single oriented, increase.
Looking at the graph makes me wonder what part of the men and women living alone, and other non-family, were formerly part of a marriage or other family setting (not counting their childhood).
The problem with the graph is that the non-uniform spacing over time obscures facts that are very relevant to interpreting the data. For example, if you look only at the "married couples without children" bar, it has remained basically constant since 1970. If you take the "married with children" graph by decade and drop the intervening ticks, you get:
There's a large decline here, but notice that over half of the decline happened during the 1970s, and then the rate of decline slowed significantly before accelerating again between 2010-2012.
When I look at the data, the conclusion I draw isn't "Most Americans are single, and that's changing the economy", it's "In bad economic times, young Americans delay making major life commitments like marriage and children, and the 1970s and 2010 are marked by significant economic uncertainty." In other words, BusinessWeek has yet again confused causation and correlation. But the latter conclusion is something that's already been widely reported and shouldn't come as news to anyone who has any 20-something friends.
I wonder how much of the increase in men living alone can be explained by economic migration? It might have become more common for a man to move abroad to find work and send money back to their family living back home.
Wow, the comment section under the BW article is filled with really pathetic, misogynistic comments. It's no wonder why so many of them commenting there are single.
Anything other than the bottom value is hard to figure out what the trend is over time, showing they add up to 100% is kind of a waste. The x axis ticks are not equidistant (10, 5, and 2 years between columns)..
Thanks for posting the interesting article. Wish the graph was better.