Newsweek's Jon Alter has joined the chorus of criticism of Hillary Clinton's campaign, calling it "one of the worst run campaigns in modern political history." But how do we know that? Mostly because she seems to losing the race for the nomination. I want to believe Hillary's campaign is not well run -- that's certainly been my personal view for the last month or two -- but we should be cautious about indulging that impulse. Virtually every losing campaign is described as badly run, particularly when the candidate once led in the polls. Hillary-bashers like me can point to various anecdotes about mismanagement and infighting, but losing causes those sorts of anecdotes to leak and become the focus of media coverage.
In a general election, the same sort of inferential problem applies -- everyone thinks Michael Dukakis ran a terrible campaign in 1988, for instance, but the state of the economy suggested that George H.W. Bush would win. More generally, the literature on projecting presidential outcomes suggests that campaigns and/or candidates may only matter on the margin.
So how can you objectively determine what a good campaign is? In a general election, you could look at how the campaign did relative to the projected outcome based on the state of the economy, though this puts a lot of faith in your model of the presidential vote. For instance, Bill Clinton outperformed the projected outcome in 1996 as estimated by the "bread and peace" model of Douglas Hibbs. Given that many experts also think he ran a good campaign, we might have more confidence in that judgment. By contrast, Michael Dukakis did almost exactly as well as the model predicts, suggesting that he was not really the problem.
The problem in Hillary's case is that we have no equivalent models for primary elections, so it's not at all clear how to project outcomes and measure possible effects of campaign management or candidate performance. In other words, beware of 20/20 hindsight on the quality of Hillary's campaign.
There's quite a bit of empirical evidence out there.
Not filing a full set of delegates in Pennsylvania is a mistake
http://atrios.blogspot.com/2008_02_17_archive.html
This NYT piece is somewhat about impressions, but notice the quotation from Amy Fried about the weakness of Clinton's internet presence. Also, this article and others point out that Obama had/has more offices in many states:
http://www.nytimes.com/2008/02/14/us/politics/14clinton.html
Emphasizing TV ads over the "ground game" isn't necessarily a mistake (what if all those office in Idaho failed?), but the Clinton campaign had the resources to at least match Obama in these places.
Having said that, spending $1200 on Dunkin' Donuts isn't necessarily a mistake. People love donuts. Maybe it was effective.
Posted by: JT | February 25, 2008 at 11:31 AM