Did Voters Give Republicans a Mandate for More War?

Everybody knows that the recent election was all about the economy, right?


Nobody would claim that American voters just gave Republicans a mandate for more war, would they?

Here we go:

 

Republican Senator John McCain said on Tuesday in the wake of big Republican victories in Congress that he hopes President Barack Obama will take a fresh look at U.S. war policy in Afghanistan. McCain won re-election to his Arizona Senate seat by a large margin, ensuring he will retain have a strong voice on the Senate Armed Services Committee as its ranking Republican member.

In an interview, McCain told Reuters he was looking forward to a December review the Obama administration is preparing to give an update on the U.S. troop increase Obama ordered a year ago to try to repulse a strengthened Taliban.

McCain, who is expected to visit Afghanistan soon, said he would like to see a change in Obama's decision to begin withdrawing some U.S. troops from Afghanistan next August.

The world would be a better place if one could just ignore things like this. But as Reuters points out, McCain is ranking member on Senate Armed Services, in addition to being the 2008 Republican presidential nominee. Two axioms of politics in America are: 1) you can't ignore a dangerous political claim, just because it's nonsensical, and 2) you can't wait for a nonsensical and dangerous political claim to gain momentum before moving to quash it, because it's like a highly infectious disease: you have to stamp it out immediately before it takes root in the population.

So, donning my Captain Obvious superhero costume, allow me to try to bring some evidence to bear on Senator McCain's apparent claim that the recent election was a mandate to extend the presence of US troops in Afghanistan.

We all know that Afghanistan was not a major issue in the election, because the press told us so. But maybe you could argue that voters tended to support pro-war candidates, even if the war was not a prominent issue.

On July 1, there was a key indicator vote in the House of Representatives on the war in Afghanistan, when 162 Representatives voted in favor of the McGovern-Obey-Jones amendment requiring the President to establish a timetable for the withdrawal of US troops from Afghanistan. The amendment would have required not just a timetable to begin the withdrawal of troops, which the President has announced, but a timetable to complete the withdrawal of troops, which the President has not announced. Thus, the amendment was significantly sharper than the policy that McCain opposes; some Democrats who support the President's timetable to begin withdrawing troops voted no on the McGovern amendment. So, if any voters wanted to send Washington a message that they want more war, they definitely should have had it in for anyone who voted for the McGovern Amendment.

60% of the House Democratic caucus voted for the McGovern amendment; 40% of House Democrats voted no. So, if the recent election was a mandate for war, it should have punished the McGovern Democrats more than the anti-McGovern Democrats.

The numbers do not bear this out.

Of the 256 Democratic seats in the last House, 119 were represented by Democratic incumbents who ran for re-election and were judged to have "competitive" races (i.e., not "safe"); 137 were not.

Of these 119 "vulnerable" Democratic incumbents running for re-election, 51 were defeated.

Among the 119 vulnerable Democratic incumbents, 53 voted for the McGovern amendment and 66 voted against it. The remaining 100 Democrats who voted for the McGovern amendment were from the 137 remainder: not Democratic incumbents seeking reelection judged to have competitive races.

Among the 53 vulnerable Democratic incumbents seeking re-election who voted for the McGovern Amendment, 12 were defeated and 41 were re-elected.

Among the remaining 66 vulnerable Democratic incumbents seeking re-election who voted against the McGovern Amendment, 39 were defeated and 27 were re-elected.

Among the 68 vulnerable Democratic incumbents who were re-elected, 41 voted for the McGovern amendment and 27 did not.

Among the 51 vulnerable Democratic incumbents who were defeated, 12 voted for the McGovern amendment and 39 did not.

So, if you were a vulnerable Democratic incumbent running for re-election who voted in favor of the McGovern Amendment, your odds of being re-elected were 77%. If you were a vulnerable Democratic incumbent running for re-election who voted against the McGovern Amendment, your odds of being re-elected were 41%

Now, there is a obviously a positive correlation between how Democratic your district is and supporting the McGovern amendment; and a negative correlation between how Democratic your district is and whether or not you are defeated by a Republican. A reasonable guess is that the fact that McGovern Democrats were more likely to be re-elected mostly reflects the fact that McGovern Democrats, on average, represent more Democratic districts.

This is a testable proposition.

If you run a simple linear regression of election outcome versus McGovern vote for the 119 vulnerable Democratic incumbents, you indeed find that McGovern Democrats were less likely to be defeated, and the result is statistically significant.

But if you run another linear regression including the Cook Partisan Voting Index for the district as an independent variable along with the McGovern vote, you find that the Cook PVI captures the lion's share of the explanation.

In the new regression, it's still the case that for a given PVI, a McGovern Democrat was more likely to be re-elected, but the result is no longer statistically significant. So, based on the standard statistical test, we can't conclude that, all other things being equal, voting for the McGovern amendment made a Democrat less likely to be defeated. But the "preponderance of the evidence" - the majority of probability - suggests that it did, although the effect was weak compared to the partisan composition of the district. Certainly, there is no evidence for the McCain claim; McGovern Democrats were not more likely to be defeated.

Of the new 184 member Democratic caucus, about 141 are either Democrats who voted for the McGovern Amendment, or Democrats who replaced Democrats who voted for the McGovern Amendment. [Two Democrats replaced Republicans: Colleen Hanabusa (Hawaii 1) and Cedric Richmond (Louisiana 2), likely McGovern supporters, I would think.]

This suggests that while McGovern supporters comprised about 60% of the outgoing Democratic caucus, they will comprise about 75% of the new Democratic Caucus. This roughly approximates the views of Democrats in the country at large: around four-fifths support a timetable for withdrawal.

The 12 defeated McGovern incumbents were:

FL-8 Alan Grayson IL-17 Phil Hare MI-7 Mark Schauer MN-8 James Oberstar NH-1 Carol Shea-Porter NY-19 John Hall OH-6 Charlie Wilson OH-15 Mary Jo Kilroy PA-3 Kathy Dahlkemper PA-11 Paul Kanjorski VA-5 Tom Perriello WI-8 Steve Kagen

If one of these 12 is your current representative, please send them a thank you note for voting against the indefinite continuation of the war in Afghanistan.

As for John McCain, maybe when he retires he can get a job writing editorials for the Washington Post.

In the new regression, it's still the case that for a given PVI, a McGovern Democrat was more likely to be re-elected, but the result is no longer statistically significant. So, based on the standard statistical test, we can't conclude that

Post new comment

The content of this field is kept private and will not be shown publicly.
  • No HTML tags allowed
  • Lines and paragraphs break automatically.

More information about formatting options

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.
Image CAPTCHA
Enter the characters shown in the image.