In the wake of the Brown Revolution, The New Republic has an article by John Judis about whether President Obama can successfully pivot to the center. Judis compares Clinton post 1994 to Carter post 1978, both Democratic presidents who receiving stunning mid-term rebukes. In Carter's case, he failed and lost re-election in 1980; Clinton, of course, succeeded. Obama has not exactly had a mid-term "rebuke", but the Massachusetts message was a shot across his bow.
It's an excellent article and well worth reading. But Judis plays some games with graphs near the end, where he compares Obama's disapproval rating to the unemployment rate and finds them closely correlated. Freely throwing cautions about correlation and causation to the wind, he then concludes that the rising unemployment rate is one cause behind Obama's disapproval. To show the problems with this approach, I produced a similar graph:
Here the other series is listed as "Mystery" so that you won't immediately know what it represents. Other than a little break early on, this is just as good a correlate as Judis'. (In fact, the mathematical correlations are nearly identical.)
OK, enough foreplay. Here's the mystery series: it's the S&P 500. So following Judis' logic, the rising S&P 500 has caused rising disapproval with Obama. Anyone believe that? Yeah, me neither.