Visual Rhetoric – Some feedback for MSNBC


Over the years, MSNBC has been using and reusing a particular visual rhetorical style for presenting job growth numbers since Obama’s taken office. And it’s not bad for presenting good data consistently over time, for keeping the range of values within a reasonable scale. All good things, but what it lacks is punch.

The graph appears to be created by a blogger who’s associated with the Rachel Maddow show. Here’s a recent form of the graph and the ensuing commentary.

The graph is about jobs created over the time scale of the chart: From January 2008 (the year before Obama took office) through January 2009 (when Obama took office) to today. What you see charted here is the rate of change in the number of jobs created, not the total number of jobs created over the time. So you can see that in the year before Obama took office we were losing jobs and then through 2009, the loss slowly diminished and by 2011, we were steadily growing.

What the graph DOESN’T show you, and I think frustratingly so, is the total number of jobs created over time. For that you’d have to add each new bar to the previous amount, successively. And granted, this would make the vertical scale kind of ludicrous, but I think that should be the point. The point is to be convincing about the graph, to show the steep rise in jobs created, not to be moderate and to show consistent job growth. I think the consistency is confusing to some folks who aren’t seeing the graph as a portrayal of rate of change.

I am reminded of an annual report I saw during the worst of the recent recession for a 529 college fund I’ve been contributing to. The managers of the fund chose to portray the loss of value in the fund over time as apparent growth by switching the years around on the time axis so that folks not looking carefully would see the upward trend and hopefully not see that the years were reversed on the X axis of the graph.

I don’t think that numbers should be manipulated for impact like that, but I do think that graphs should show off, and forcefully, the main conceit. And the problem with a graph showing rate of change is that it sort of averages out the accomplishments instead of highlighting them, as it should.

For more look at graphs of derivative equations versus original formulas. For example, this graph generated by WolframAlpha.

Or feel free to ask a question!

, , ,