Hillary Clinton has escaped indictment, but the FBI’s characterization of her as having been "extremely careless" in using multiple "personal servers" to send "Top Secret" emails—and as perhaps not having been "sophisticated enough" to understand classified markings while serving as U.S. secretary of state—appears to have cost her public support. Indeed, on the cusp of the national political conventions, the presidential election now looks like essentially a dead heat.
In something of a reversal from earlier polling, national polling now favors Clinton while state-by-state polling favors Donald Trump. Only one national poll has been taken (mostly) since the FBI announced its findings. That poll, from McClatchy/Maris, finds Clinton up 3 points over Donald Trump (42 to 39 percent) and up 5 points (40 to 35 percent) when Libertarian candidate Gary Johnson (10 percent) and Green candidate Jill Stein (5 percent) are included.
State-by-state polling (per RealClearPolitics), meanwhile, offers better news for Trump. As Shoshana Weissmann reports, in polling done about half before and half after the FBI released its findings, Quinnipiac finds Trump up 3 points in Florida (42 to 39 percent) and up 5 points (41 to 36 percent) if Johnson (7 percent) and Stein (4 percent) are included. JMC Analytics, in polling taken entirely after the FBI's announcement, also finds Trump up 5 points in Florida (47 to 42 percent) in a four-way race (with Johnson at 2 percent and Stein at 1).
Meanwhile, Monmouth finds Trump up 2 points in Iowa (44 to 42 percent) in a four-way race (with Johnson at 6 points and Stein at 1), in polling taken entirely after the FBI's announcement, while it finds Clinton up 4 points in Nevada (45 to 41 percent) in a three-way race (with Johnson at 5 points). In polling also taken entirely after the FBI's announcement, Harper finds Clinton up 7 points in Colorado (45 to 38 percent). In Pennsylvania, Quinnipiac finds Trump up 2 points (43 to 41 percent) and up 6 points (40 to 34 percent) if Johnson (9 percent) and Stein (3 percent) are included. In Ohio, Quinnipiac finds a tie (41 percent apiece), with Trump up 1 point (37 to 36 percent) if Johnson (7 percent) and Stein (6 percent) are included.
To beat Clinton, Trump would have to fare about 5 percentage points better than Mitt Romney in terms of the margin of vote. (Romney lost to President Obama by 4 points but would have lost the electoral vote even if he'd narrowly won the popular vote, given his terrible performance in swing states.) Romney lost by 1 point in Florida, 3 points in Ohio, 5 points Pennsylvania and Colorado, 6 points in Iowa, and 7 points in Nevada. Aside from the one poll in Colorado, Trump is now faring 3 to 9 points better in these state polls, in terms of the margin of vote, than Romney fared in the same states during the 2012 election.
It will be interesting to see how or if the conventions affect what now looks to be a race with no clear leader.
Update: Polls subsequently released by NBC/WSJ/Marist, Gravis, and Marquette find similar results in Ohio (where NBC/WSJ/Marist finds a tie in a two-way race and a 3-point Clinton lead in a four-way race) and in Iowa (where NBC/WSJ/Marist finds Clinton up 3 points in two-way race and tied with Trump in a four-way race, and where Gravis finds Clinton up 2 points in both a two-way and a four-way race). They find wildly different results in Pennsylvania (where NBC/WSJ/Marist finds Clinton up 9 points in a two-way race and up 8 points in a four-way race, more than 10 points off in each case from Quinnipiac's results). They provide further evidence that Trump seems to be struggling badly in Colorado (where Monmouth finds Clinton up 13 points in a four-way race). And they find Clinton leading in Wisconsin (where Marquette finds Clinton up 4 points in a two-way race and up 6 points in a four-way race), a state in which Obama beat Romney by 7 points. The combination of these polls suggests that Clinton likely has a narrow lead going into the conventions.