Enos & Fowler's Map of Overlapping Media Markets in Battleground & Safe States
The title of a recent WaPo article asks a pertinent question: “How many votes will Trump give up by not running a professional campaign?” The whole article is well worth reading, but the article’s conclusion can be summarized in the two words that complete the title: “A lot.” Still, exactly how much is “a lot”? That’s the million dollar question. This diary will attempt to get a little closer to an answer to that question—though we’ll need to wait until election day to know the exact impact of the vast difference the the two campaigns field operations.
I looked at the essay by political scientists Ryan Enos and Anthony Fowler on GOTV referenced in the WaPo piece. Their basic method was to look at segments of adjoining states with a shared TV market, but with one state a battleground and the other not. TV advertising would be the same, but one would see heavy GOTV and the other wouldn’t. The map at the top of this diary shows the areas studied. The areas studied are enclosed in black lines. The darker the area is colored, the greater the likely GOTV effort. One can clearly make out from the map that, for instance, hotly contested NH would likely get lots of GOTV effort, but adjacent VT, MA and ME (all safe D) would not. Similarly, northern NV would get intensive GOTV efforts, but overlapping media markets in CA & OR (safe D) and Utah (safe R) would not. And in fact the darkly shaded areas—the likely targets of GOTV— showed significantly increased turnout over the light gray areas.
The gist of the study is neatly summarized by WaPo:
They show that the effect of the 2012 presidential campaign on voter turnout was quite large, about 7-8 points overall.
Now 7-8 points sounds like an awful lot. But what the study claims is not that one side beats the other by 7-8%, but that EACH side, if they have equal operations, turns out an extra 7-8 by a serious GOTV effort. If so, that would suggest that an awful lot of GOTV effort could be considered defensive—you’ve got to at least equal what the other side is doing, or you’re going to get buried.
So the 1-2% margin we often see mentioned as a GOTV effect might be the difference between an outstanding GOTV effort vs. just a garden variety effort: one side is getting out only an extra 6% and the other is getting out an extra 7% or 8%.
On the face of it, this makes a lot of sense—could all those phone calls and pavement pounding and personal visits really only amount to a measly 1%? It seems intuitively right that this 1% margin would actually the net effect of two operations at work, one just a tad better than the other. 7 to 8% could be considered the gross effect. All of this means kudos to the people working hard in field operations—you’re adding 7-8% to total turnout, and without it, we wouldn’t have a prayer. BTW, Enos & Fowler credit both Obama & Romney with having effective GOTV efforts, so if Obama had an edge, it might well have been in the 1 or 2% range.
But if Enos & Fowler’s analysis of the gross impact of GOTV efforts is correct, then this might suggest that Trump’s neglect of GOTV will have a much larger impact than many people think. Of course, Hillary probably won’t outdo the polls by 7-8%, since the RNC and state party machines have their own GOTV efforts going.
But we might reasonably expect a much bigger effect in the battleground states this year, based on Clinton’s vastly superior operation, than the traditional estimate of 1-2% for a better field operation. Of course, we won’t really know until election day.
But however you slice it, we can regard the present election as a living, breathing experiment in the impact of GOTV. This makes me a bit more optimistic about the possibility of flipping some Senate and House seats that might otherwise might be out of reach. So let’s all keep working hard, not just for that slim 1% but for a potentially much larger slice of the pie!
