I’ve been thoroughly disappointed all week about the result of Pennsylvania’s primary, but finally sucked it up this morning and took a look at the New York Times exit polls to try to get an understanding of Pennsylvania’s demographics and what caused Clinton’s victory (and what, if anything, Obama needs to do differently there to win this fall).
In doing so, I stumbled upon some statistics that strongly suggest that McCain supporters voting for Clinton could have caused a three percentage-point swing in Clinton’s favor. While this would obviously not be enough to change the victor in the Pennsylvania primary, it appears that this phenomenon (first observed in the Texas primary) is having more than a trivial effect on the overall contest. My analysis is below the fold.
When reviewing the exit polls, I noticed that there were a number of Clinton voters who, in a hypothetical matchup between Clinton and McCain, would either vote for McCain or stay at home:
-- 11% of the primary voters (or 253,733) said they would vote for McCain over Clinton in the general election; of these, 29% (or 73,583) voted for Clinton in the primary.
-- 6% of the primary voters (or 138,400) said they would not vote in a hypothetical Clinton versus McCain matchup; of these, 6% (or 8,304) voted for Clinton in the primary.
Thus, overall, based upon exit polling, 81,887 voters -- 3.6% of the total votes cast and 6.5% of Clinton's voters -- voted for Clinton in the primary but would vote for McCain or stay home in the general election.
If you were to remove these 81,887 votes from Clinton's total, Clinton would have accumulated 1,178,557 votes, or 53.0%, while Obama would have received 1,046,220 votes, or 47.0%. This would have the effect of reducing Clinton's margin of victory from 9.2% to 6.0%.
This raises a significant question: Who are these 81,887 people, and why would they vote for a candidate in the primary election but not in the general election? Were they wolves in sheep’s clothing, so to speak?
While I hate the thought that Limbaugh’s call for Republicans to vote for Hillary affected the Pennsylvania primary, it does appear that a number of Republicans did attempt to manipulate the process (as was explored yesterday by Moesee). The Washington Post reported that 164,000 Pennsylvania voters switched their registrations from independent or Republican to Democratic, and speculated that it could have been due to the Limbaugh effect, but interviewed several people who said they were genuinely switching because they supported Clinton or Obama. On the other hand, ABC News gave anecdotal accounts of Republicans registering as Democrats to throw a wrench in Obama’s campaign:
Republican crossovers are a key factor in record-level voter registrations for the Pennsylvania Democratic presidential primary, officials say.
* * *
While many of the new Democrats appear to be moderates or independents who simply want to be a part of the process, county voter registration officials in central Pennsylvania told ABCNews.com that many new registrants spoke openly about changing their party affiliation to give McCain "a better shot in November."
Officials in Perry and Northumberland counties in central Pennsylvania told ABCNews.com that quite a few new registrants said they were switching to help the Republican party in the fall. Both counties are historically conservative, having voted for Bush in 2004 in wide margins, but the number of voters changing their party affiliation to Democrat this year is proportionally large, said the officials.
Other local newspapers offered similar anecdotal evidence. For instance, in Adams County, the Evening Sun, in an article entitled "Voters Switch Parties to Help McCain,", reported:
Sandra Reed of Gettysburg has been a Republican since she was old enough to vote.
But, Tuesday, she and her husband, Vernon, went to the Adams County Courthouse and became Democrats.
"We were registered Republicans, and we will always be Republicans, but we want to help Hillary get the No. 1 position for the Democrats," said Reed, 70. "So, we are switching for the primary to vote for Hillary, then we will switch back and vote for McCain."
* * *
They got the idea to influence the Democratic race from conservative radio talk-show host Rush Limbaugh, she said. Limbaugh recommended his listeners temporarily switch party affiliation to give Clinton a boost, and vote out Obama.
Monica Dutko, Adams County elections director, has seen a steady stream of Republicans switching over to the Democratic Party. While she does not ask why they're switching, some have told her that they're switching at the suggestion of Limbaugh.
* * *
In York County, Director of Elections Nikki Suchanic is seeing a similar trend. She said many are inquiring on switching parties, and it may be because Pennsylvania can be an influential state in the Democratic race, but not the Republican race, which McCain sealed last week.
I know that my analysis here is far from scientific, and would be interested in hearing others’ thoughts (even if it’s to tell me to take off my tinfoil hat for awhile this weekend). While there are only a few primaries remaining, Indiana is an open primary, and I fear that this trend of Republicans manipulating our primary contests could be even more prevalent there. Thanks for bearing with my math!
Update: A fair criticism of this diary was that I did not crunch the numbers the other way -- I did so in a comment, but wanted to include it in the substantive diary:
(again, going by the NY Times exit poll numbers):
-- 15% (or 346,000) voters in a hypothetical Obama-McCain matchup would vote for McCain; of these, 10% (or 34,600) supported Obama in the primary.
-- 10% (or 230,666) voters in a hypothetical Obama-McCain matchup would not vote in the general election; of these, 2% (or 4,613) supported Obama in the primary.
Thus, a total of 39,213 voters supported Obama in the primary but would not support him in the general, as compared to 81,887 for Clinton. If you deducted Obama's 39,213 and deducted Clinton's 81,887, the overall vote would be 1,178,557 votes, or 53.9%, for Clinton, and while Obama would have received 1,007,007 votes, or 46.1% -- a 7.8% margin.