But when Taylor leaps from that correlation to saying that what we need for economic recovery is to “lighten up on the anti-business sentiment coming out of Washington,” I wonder what is going on in his head.
. . . I have no dog in the fight, and can sit back and watch with growing amusement.
Mark Thoma supplies links, so I don't have to. But it takes a few click-throughs to get to this one by Noah Smith, which, because of its insights, really should not be missed.
Another insight worth noting is this one from a much younger John Taylor, quoted by Justin Wolfers.
Behind the use of time-series estimation is the assumption that the structure and coefficients to be estimated remain stable over the sample period… If this assumption is in fact not correct, the estimated function would have little use either as an explanation of the causes of murder or of the policy implications of changing the value of an exogenous variable in the structure.
I think about and use time series data a lot - in fact as recently as yesterday. So this is something worth keeping in mind when any of us look at those kinds of data series. Though I'm usually looking for what changed, rather than some underlying determinant.
There are other lessons here about cherry picking, data analysis, and data mining that should make you cautious of any graphic presentation - no matter how convincing it seems to be.
One of my father's catch phrases was, "Figures might not lie, but liars sure know how to figure." He was born in April, but was certainly no fool.
Update: Just to be clear, I'm not calling Taylor or anyone else a liar. The quote above from my dad, in red, is a general caution, not an indictment.
Update 2: Karl Smith weighs in, and finds: "In short, in either of these basic comparisons I just don’t see a lot of there there."