FORTUNE — Gautam Mukunda has written a new book that tries to answer two vexing questions: Does it matter which leaders we choose? And if so, can we make rational choices? In Indispensable: When Leaders Really Matter, the Harvard Business School professor argues that “unfiltered leaders” — those who haven’t had much relevant experience — will likely have a greater impact than filtered leaders. What’s harder to predict is whether that impact will be good or bad.
Mukunda’s take on the HBS grad who’s currently running for president? “Romney’s business experience tells us little about what he will do in office, and what he learned during it will be of little help to him.” Electing Romney, Mukunda says, would be a “tremendous gamble.” (The following interview was edited for space and clarity.)
What inspired you to write this book?
In general, academics who study leadership have concluded that it’s just not that important. There’s some good statistical evidence for that, there are good theoretical arguments, it’s a persuasive set of beliefs. But the problem is that no one outside academia believes that for a second. Just no one. Whenever I see something like that, I assume that something interesting is going on. This is probably the single greatest variance in beliefs about the world between social scientists and practitioners.
When academics say that leaders don’t matter that much, what they’re thinking is, okay, if that person had not been there, the person who would have had that job instead would have done roughly the same thing. Whereas people outside of academia tend to think that leadership and management matter a lot. So leadership and management can matter, even if leaders or managers might not, because they’re replaceable.
But we can all think of countless examples where individual leaders were not replaceable. I don’t think there’s anyone who thinks that if Steve Jobs had not run Apple, Apple (AAPL) would still be a $600 billion company.
Are high-impact leaders always the best choice?
If our definition of leader-impact is that this person does things that other people in the same situation would not do, what do we know about that kind of decision? Think about what happens when you buy a stock that everybody else says is going to tank. Most people will say you’re crazy. If you’re right, and it shoots through the ceiling, you’re a genius. But most of the time when everybody is telling you, ‘Gee this is an awful idea,’ it is actually an awful idea.
The peril of high-impact leaders is precisely this: They’re the people who do the things that no one else would do. Sometimes they look great, and they’re Abraham Lincoln. Sometimes they’re awful, and they’re Warren Harding.
So how do we identify high-impact leaders?
That’s where the filtered/unfiltered idea comes in. Look at how GE chooses a leader. They don’t hire outsiders. They bring in people at low levels and they spend 20, 25 years evaluating them. By the time someone is in contention to be CEO, GE (GE) knows everything there is to know about this person. All those years—with their peers, with their superiors, with their subordinates—watching them perform in every conceivable circumstance. At the end of the day, GE’s got five finalists for the job, and they pick one. All of the finalists have been through this process. How different could the four who weren’t chosen possibly be?
But that’s not how we choose presidents.
Right. A disproportionate percentage of both the most effective and the least effective presidents have been unfiltered.
Romney would be in that category?
Yes. He is an extraordinarily unfiltered leader. He has less relevant experience than any Republican president has ever had before entering the Oval Office, and less than any president of any party except Woodrow Wilson. Grover Cleveland had slightly more experience in lower offices, he’s the other comparison.
Doesn’t Romney’s business experience count for something?
We have a lot of research that tells us experience doesn’t transfer from one context to another very well. We know that when you bring stars from one company to another, even in the same industry, they usually don’t stay stars. That’s true even when you’re moving within an industry. So think about how much more true it’s going to be when you’re moving from finance to politics.
There’s no question that Mitt Romney is filtered as a private equity executive. If you wanted to hire someone to run a private equity fund, Mitt Romney’s got to be on the really, really short list of people you would pick. But that sort of evaluation doesn’t tell you a whole lot about his ability to succeed in politics.
Romney argues that because of the economic fix we’re in, his business experience is uniquely relevant in this election. You don’t buy that?
I don’t. I think his business experience would tell you lots of things. The job of someone in a private equity fund is to maximize return on investors’ capital. He’s incredibly good at that. But the job of president is not to maximize return on investors’ capital. The president of the United States has 300 million people he has to benefit, and they have lots of different interests. Can he form coalitions with people who have no reason to work with him? Can he rally the public? Can he negotiate keeping in mind that he serves many, many different constituencies?
Are you saying that Romney’s business experience would be a handicap, not an advantage?
It’s both. The business experience tells us he’s incredibly smart, he’s hardworking, he’s analytical. He’s able to solve problems creatively and intelligently. These are all good things in a president. They should give us some confidence that we’re not picking someone at random who has never done anything that would make you feel they’re capable of being president. But at the same time it’s a profoundly different set of skills.
As with any other unfiltered leader, Romney would be an extraordinarily high-variance choice. He is likely to do many things that no one else would do. Will that be better or worse? I will say with some confidence that he will make mistakes that no one else would make. We’ve already seen this in the campaign. He’s made mistakes that are almost inexplicable, like saying ‘A bunch of my friends own NASCAR teams.’ That seems superficial. And it is. Except that most people who spent their careers in politics would never say something like that. When you make mistakes that easy, you’re going to make other mistakes.
President Obama: Filtered or unfiltered?
Compared to Romney, he’d spent more time in politics prior to running for president. But when he was elected he was clearly massively unfiltered. This was a leap into the unknown.
Who will you vote for in November?
I’m gong to be voting for Obama.
Is that a choice informed by your research?
It’s partly informed by the research, in the sense that Obama is a potentially high-impact leader whom we’ve seen for the last four years. There are things he’s done that I disagree with. But I believe the areas in which unfiltered leaders are most likely to fail—foreign policy, for example—he’s proven he’s not going to fail. He’s not completely incompetent, he’s not crazy, he’s able to do the blocking and tackling of the job. With Romney I say we have the potential upside of the unfiltered leader but we also have the potential downside, and we don’t know what’s going to happen.