Data Viz Done Right

November 27, 2009

Sometimes a dual axis is not a dual axis

No comments
I learned something after my last post, thanks to my friend and co-visual analysis geek (or is it enthusiast??) Joe Mako. The title from the graph I was referring to was "Fed Funds Rate vs. 30-Year Fixed." That right there should have told me the graph was a comparison, but the fact that there was an axis range on both sides of the graph, led me immediately to assume it was a dual axis chart. We all know why you don't assume...

Joe point out to me that the subtitle is "Interest Rate Differential Since 2000." That's the key to the chart...differential. Maybe it says something about the chart that I didn't notice the subtitle or that my eyes were drawn to the ranges on both axes, but I made a mistake. Phew, that felt good to say.

Joe recommended recreating the graph that I had previously posted with a single axis range, since the ranges were so close already.

I also wanted to look at the differential, since the author's point was to show that the differential between the Fed funds rate and the 30-year fixed mortgage rate was that the Fed funds rate only influences the 30-year fixed rate. If the Fed established mortgage rates, then the chart would be completely linear, which it clearly is not.

I chose the color red since the farther from the zero axis, the less influence the Fed has on mortgage rates. The darker the red, the less the influence.

I also changed the title so that it would be more clear what the chart was comparing. The author of the article titled the chart "Fed Fund Rate vs. 30-Year Fixed." When I recreated the chart, I simply took the Fed rate and subtracted the 30-year fixed rate, but that made the chart a mirror image of the author's, meaning that he had the title backwards in my opinion, thus the title I arrived at.

The bottom line is that I agree with Dan Green's evidence...the two rates are NOT strongly correlated.

No comments

Post a Comment