The VP Choice Effect
I've already looked in the past at the presidential nominee's home field advantage and concluded it was worth somewhere between five and seven percentage points. That is, presidential nominees tend to do about five to seven points better in their home state than they do nationally.
It seems logical to assume that the VP nomination has a similar effect; all other things being equal (which of course they aren't), a state that has a VP nominee should tend to vote a little more for that ticket than they might otherwise. We would assume that the effect would be smaller than for the presidential nominees.
So I put together a little spreadsheet that analyzes the past impacts of VP nominees. I'll illustrate with John Edwards, John Kerry's running mate in 2004. Edwards was from North Carolina, which Bush took by 12.44 percentage points in 2004. By contrast, Bush won North Carolina by 12.83 percentage points in 2000, so Edwards improved things by about 0.39 percentage points.
But that's not the whole story. Remember, Bush actually lost the popular vote nationally by 0.51 percentage points in 2000, and won by 2.46 percentage points in 2004, so there was a national trend for Bush to pick up 2.97 percentage points. Thus, the indicated improvement for the Democrats from having Edwards on the ticket appears to be about 3.36 percentage points.
Of course, there are lots of other things going on in the race, so it's impossible to say that was definitely the Silky Pony effect, but it appears to fit in well with prior years. I went back for the last 16 veep nominees. I excluded sitting vice presidents for two reasons. First, they only run with sitting presidents, and it seems far more likely that voters vote on the performance of that president rather than for their local boy as VP. And second, we don't face that situation this year.
Year VP Nom State Result Prior Trend Effect
2004 Edwards NC -12.44% -12.83% -3.0% 3.36%
2000 Cheney WY 39.79% 12.97% 8.0% 18.82%
2000 Lbrman CT 17.57% 18.14% -8.0% 7.43%
1996 Kemp NY -28.86% -15.85% -3.0% -10.06%
1992 Gore TN 4.65% -16.34% 13.3% 7.71%
1988 Quayle IN 20.15% 23.99% -10.5% 6.65%
1988 Bentsen TX -12.60% -27.50% 10.5% 4.40%
1984 Ferraro NY -8.01% -2.67% -8.5% 3.13%
1980 Bush TX 13.86% -3.17% 11.8% 5.23%
1976 Mondale MN 12.88% -5.51% 25.2% -6.82%
1976 Dole KS 7.50% 38.16% -25.2% -5.45%
1972 Shriver MA 8.97% 30.12% -22.5% 1.30%
1968 Muskie ME 22.23% 37.68% -23.3% 7.83%
1968 Agnew MD -1.55% -30.96% 23.3% 6.13%
1964 Hmphrey MN 27.76% 1.38% 22.3% 4.07%
1964 Miller NY -37.25% -5.26% -22.3% -9.68%
As you can see, 12 of the 16 nominees did improve their ticket's anticipated performance; the only ones who failed to do so were Kemp in 1996, both Dole and Mondale in 1976, and Bill Miller in 1964. Cheney in 2000 appears to have improved his ticket the most in his home state, but it should be noted that only Al Gore in 1992 appears to have swung his home state over into the winning column. And even there caution is in order; Bill Clinton was also a southerner and that may have had more to do with winning Tennessee than the presence of Gore. Overall the average VP appears to have resulted in a 2.75 percentage point improvement in his home state over what the ticket would have done without him (or her, in Ferraro's case). This seems a reasonable estimate. It's about half the presidential home field advantage.
Implications for 2008? Here's where things get really complicated. We cannot assume that McCain will carry the country by the same 2.46 percentage points that Bush did in 2004. And we cannot assume that the Democrats will win big as some of the polls are indicating. It seems safest to start both parties at even, by deducting 2.46 percentage points from the Republican net percentage in each state. This tips three states that Bush carried in 2004 into the Democrats' column: Iowa, New Mexico, and Ohio. Obviously Ohio is the most important of those three; if McCain carries all the states Bush did except New Mexico and Iowa he still wins (barely) with 274 electoral college votes of the 270 needed. Thus it would seem that an Ohio politician like former Senator Rob Portman would be a logical choice.
Of course, there are lots of other variables in the race. It is well-reported that if Obama wins the Democratic nomination that the state of New Jersey may come into play, which might indicate a prominent NJ politician like Christie Todd Whitman could swing the Garden State to the GOP. That would have the added effect of capitalizing on the resentment some women may feel with Hillary not getting the nod in the Democratic Convention, although it would not help McCain with the conservatives. But McCain could lose Ohio, New Mexico and Iowa and still win by picking up New Jersey.
For the Democrats, a lot depends on whom the nominee is, but Florida and Colorado would both seem to be within reach. Colorado's governor is Bill Ritter; the only problem is that he has even less experience in elective office than Barack Obama, having been elected in 2006. Bill Nelson would be the obvious choice for Florida.
Labels: 2008 Candidates, Bill Nelson, Bill Ritter, Christie Todd Whitman, Rob Portman, Vice Presidents