AP
Hillary Clinton.
After nearly four decades of controversies around her and her husband, former President Bill Clinton, people have already decided whether they trust her or not, and they won't be moved by new stories about her email server or Benghazi talking points or anything else.
And while her support for the Iraq war was a negative for Democratic voters, the polls indicated that they had decided to forgive her for it.
The surprising and persistent success of the Bernie Sanders campaign showed that I was wrong: Many of Clinton's fans could be persuaded to vote for someone else.
He managed to do so in large part by focusing on two issues on which Clinton had not been previously attacked much: Wall Street, and the concern that Clinton is too cozy with it, as exemplified by her paid speeches to Goldman Sachs; and free trade, which Clinton has intermittently supported.
Tuesday night's result in Michigan was shocking, in that Sanders narrowly won when the polls said he would lose by nearly 20 points. A new CNN poll out today has Clinton leading by 30 points in next week's Ohio primary, while a Chicago Tribune poll last week had her leading by 42 in Illinois. But I'm not sure I believe either poll in light of the Michigan results.
That said, while this loss is surely disorienting for the Clinton campaign, they shouldn't panic, for a few reasons.
The first is that Sanders did not win Michigan by enough to be on pace to overtake Clinton in pledged delegates. Last month, FiveThirtyEight produced a useful chart of where Sanders needed to win, and by how much, in order to fight Clinton to a draw.
To win nationally, you have to win your most demographically favorable states by a lot, in order to offset losses in your weak regions. The calculations are fairly simple on the Democratic side because all states award their delegates proportionally.
The FiveThirtyEight calculations showed Sanders needed to win Michigan by four points, and he won only by two. But the news is worse for him than that. He's underperformed his targets in most of the voting states to date. For example, he needed to win Massachusetts by 11 points, and instead he lost by one. So, in order to make up lost ground, he needed to outperform his targets in Michigan (and everywhere else) by a significant margin. He didn't.
Thomson Reuters
Democratic U.S. presidential candidate Clinton speaks as rival Sanders listens at the Democratic U.S. presidential candidates' debate in Flint
The other reason for Clinton not to panic is that trade likely matters more as a political issue in Michigan than in any other state. If the Michigan results present the worst case scenario for how much a fight over trade can hurt her in a contest against Sanders, she will still be on pace to win a comfortable majority of delegates.
Analysts have been blown away by Clinton's reduced margin among black voters in Michigan. Exit polls suggested she won them by just a 2-to-1 split, compared to 9-to-1 in Mississippi. Is this about Clinton's super-solid support from black voters being a phenomenon specific to the South? Or is it about black Michigan voters, like white Michigan voters, being especially likely to care about trade?
We will have test cases next week: Illinois and Ohio are also Rust-Belt states, but they haven't been as hurt by global manufacturing trade as Michigan has. If Clinton manages a solid win in both states - even a win by much less than the landslide margins shown in current polls - that will be a sign the bleeding was confined to Michigan, and that she might still wrap up the nomination well before all the voting is done in June.
There is one more question for Clinton: Even if she shouldn't worry that she will fall behind Sanders for the nomination, should she worry that a Republican opponent will successfully use the same attacks Sanders has in a general election?
Facing a normal Republican, this wouldn't be a big issue. A normal Republican would be more committed to free trade than Clinton, who now says she opposes the Trans-Pacific Partnership. And a normal Republican would be hesitant to bash the banks, both because that feels anti-business and because it would offend wealthy Republican donors who work for or own the banks.
Donald Trump is not a normal Republican. He has capitalized on the same anti-trade, anti-elite sentiments as Sanders and would co-opt many of Sanders' talking points in the general election, trying to win over his fans from the primary.
But fortunately for Clinton, there is a lot about Trump to turn off Sanders voters: His support for huge tax cuts on the wealthy, his calls for the United States military to commit war crimes, his crude and offensive stereotyping of minority groups - take your pick. Plus, Trump will surely alienate some voters who normally vote for normal Republican candidates, offsetting whatever gains he might make among Sanders voters.
Still, running against Trump will require a different campaign than the textbook one the campaign would run against Ted Cruz, and one that is more focused on defending states like Michigan.
All of which is to say, Clinton has been extremely lucky. If she had faced a Democratic opponent who could capitalize on the same themes as Sanders while being more acceptable to party elites - someone like Massachusetts Sen. Elizabeth Warren or Ohio Sen. Sherrod Brown - she'd be losing. And if she faced the prospect of a populist, anti-trade Republican opponent without all of Trump's baggage, she'd be at a disadvantage in the general.
But you can't beat something with nothing, and Clinton's advantage remains that nobody she is running against is well-positioned to beat her nationally.