The Washington PostDemocracy Dies in Darkness

Hurricane Sandy was huge — but not for the 2012 presidential election

Steve Helber / AP

Today’s anniversary of Hurricane Sandy is an occasion to revisit the end of the 2012 presidential campaign, when commentators speculated about how much the hurricane and its aftermath would affect the election.  Of course, Hurricane Sandy’s toll in lives and property is more important than any impact on electoral politics.  But as I discussed at the time, natural disasters can influence elections.  So what happened when Sandy made landfall only about a week before Election Day?

As it turns out, not much.  This is the argument in Lynn Vavreck’s and my book on the campaign, The Gamble — which I’ve been writing about on and off these past few weeks.  I think four key things are important for understanding why the hurricane was largely a non-factor in the presidential race.

First, the notion that Romney had “momentum” before Sandy hit gave rise to the idea that Sandy arrested this momentum and handed the race to Obama.  But there was no evidence of momentum for Romney.

Second, the hurricane was only a temporary interruption for the campaign.  Obama canceled events on Monday, Oct. 29, when the hurricane hit.  He spent time visiting the destruction on Oct. 30-31.  Then he returned to campaigning on Thursday, Nov. 1.  Romney turned a Dayton rally into a “storm relief event” on Oct. 29 and then canceled events only on Oct. 30. He was back campaigning in Florida on Oct. 31.  Moreover, neither candidate pulled his ads.

Third, neither candidate got much more news coverage or more positive news coverage because of the hurricane.  This is especially relevant for Obama, who some presumed would get favorable coverage because he could mobilize the trappings of his office to show concern and support for the victims.  But as I noted previously, there was little evidence Obama was covered more favorably — even after Sandy.

Finally, and unsurprisingly given all of the above, the polls didn’t move much in the wake of the hurricane.  The Guardian’s Harry Enten examined the polls and found no consistent evidence of movement toward Obama after Sandy. The majority of national polls found that Obama had lost points, gained no points, or gained only 1 point after Sandy. In the state polls, Obama gained in only three of 12 states, or in 25 of the 67 instances where a pollster had polls in the field pre- and post-Sandy. There was not even a notable shift toward Obama in the states hit hardest by Sandy: New York and New Jersey.  The same was true in the YouGov polling data Vavreck and I analyzed in the most depth.

The only other political science research specifically on Hurricane Sandy is an article by Yamil Velez and David Martin of Stonybrook University.  Examining hurricane damage and vote shares at the county level, they find some evidence that the hurricane may have helped Obama, after trying to account for other relevant factors.  But they estimate that really only one swing state — Virginia — might have been affected had the hurricane never happened.  They do not believe that the hurricane ultimately affected the national result.

I have no doubt that Hurricane Sandy will live on in election folklore.  But the evidence suggests its impact on the outcome was minimal.