A question recently came up during a conversation I had with Jason Reifler: what makes someone a good political candidate? Right now, political science models focus on variables like partisanship, economic performance, etc., and don’t take the personal qualities of the candidate into account. Is is true that personal qualities don’t matter, or is the problem that we can’t measure them well? And what should we measure that can be measured? For instance, the best data we have for Congressional candidates simply measures whether they had run for public office before or not, which is pretty crude. What are we missing?

In terms of whether quality candidates matter, I’m thinking in particular of our good friend HRC. The argument I’ve made is that she is a well-defined public figure with high unfavorables, which makes her a bad candidate.

Amy Sullivan made a similar claim recently in the Washington Monthly:

[W]hile her “favorables” are good—57 percent of Americans have a positive impression of her—her negatives are disturbingly high as well. This long before an election, most voters have yet to make up their mind about a candidate. Even as close to the primaries as December 2003, 66 percent of voters didn’t know what they thought of John Kerry. That’s not the case with Clinton. While at this point in George W. Bush’s first presidential campaign, Bush also had favorable ratings around the mid-50s, an additional 30 percent of voters said they either hadn’t made up their minds about him or they didn’t know who he was. Compare that to Hillary: Only 7 percent of respondents aren’t sure what they think of her, and—not surprisingly—no one says they haven’t heard of her.

Never in American political history has a candidate faced such a decided electorate at this early a point in a presidential race. That’s a disadvantage when you consider that one of the lessons of 2004 was that once voters develop a perception about a candidate, it’s as immovable as superglue. No one who thought George W. Bush was a likable, friendly guy could be convinced that he was corrupt or misleading. And once John Kerry became identified in voters’ minds as a “flip-flopper,” no amount of arguing could change that image. It’s a problem for any candidate. For Sen. Clinton, it could be fatal. Americans know exactly what they think of her. And nearly 40 percent say they would never consider voting for her.

But there are a few potential counter-arguments:

1) Personal qualities don’t matter; the economy does. Models focused on the economic predict presidential voting very well. Candidates are just stand-ins. So Hillary will win in 2008 if the economy is bad, and lose if it’s good.
2) Hillary is polarizing, but Bush’s 2004 campaign proves that a polarizing figure can win the White House if they turn out their base (this is arguably related to #1).
3) Personal qualities don’t ultimately matter because they are socially constructed and change over time. Al Gore was seen as a Boy Scout before he was completely redefined as a liar/leftist/etc. from 1997-2000, and now his profile looks like Hillary’s. In the jargon of social science, they’re endogenous. Dick Morris has made a similar argument that “charisma” is socially constructed.
4) Hillary has a unique (but completely unproven) ability to move the Democratic Party to the center, which can help increase a party’s vote share depending on the model you use and the assumptions you make.

What do you think?