When it comes to thinking up new ideas upon which to found a business, and having the skills to turn a startup into a success, men and women begin on a level playing field. But in practice, men still heavily outnumber women in the high-risk world of entrepreneurship and venture-capital backed startups. Which leads to the obvious question: Why?

The answer, says Wharton management professor Ethan Mollick, is multifaceted, but at its core can be traced to a few fairly simple points of personality and gender tendency on the one hand, and the broadly human tendency for birds of a proverbial feather to flock together on the other. In an interview on the Knowledge at Wharton show on Wharton Business Radio on SiriusXM channel 111, Mollick spoke about a recent research paper that he coauthored with Venkat Kuppuswamy from the University of North Carolina — titled, “Humility and Hubris: Gender Differences in Serial Founding Rates” — and how not just women, but many disadvantaged groups, can overcome impediments to success. (Listen to the podcast at the top of this page.)

An edited transcript of the conversation appears below.

Knowledge at Wharton: Ethan, please explain your research to us.

Ethan Mollick: If you look overall at the chance of someone starting a company, it turns out that gender is a really strong predictor of whether or not they will become an entrepreneur. Women are less likely to be entrepreneurs than men, and this has been a big puzzle, because women are as innovative [as men and] companies run by women are as successful. So why aren’t women launching companies at the same rate?

That’s the puzzle that we were trying to solve with this research. What we did first was think about why people start companies. There’s actually a lot of research on this, and it shows that overall, entrepreneurship is kind of like buying a lottery ticket. You’ve heard the statistic that something like half of companies fail? It’s a little over that over a five-year period of time.

Most people who play the entrepreneurship game lose. So in order to be an entrepreneur, you have to be overconfident. You have to believe that you’re better than everyone around you.

In fact, overconfidence is the biggest psychological predictor of whether or not you’re going to become an entrepreneur. Having misplaced confidence in yourself and thinking you can win when other people always lose is a strong predictor of entrepreneurship. We call this kind of overconfidence classic, Greek-style hubris — the idea of unfounded self-confidence.

“Overconfidence is the biggest psychological predictor of whether or not you’re going to become an entrepreneur.”

We were thinking about this hubris result — this was work I was doing with Venkat Kuppuswamy at the University of North Carolina — and we realized that there is actually another set of research on gender that has found something across cultures and across ages called the “male hubris, female humility” effect. What it says is that women have lower levels of hubris than men — they’re less likely to be overconfident.

They’re especially less likely to make what we call the fundamental attribution error — in this case, holding the belief that when things go right, it’s all because of your genius, and when things go wrong, it’s because of luck or outside forces.

Men are much more prone to the fundamental attribution error. They’re much more likely than women to believe that their success is because of their own doing and failure is someone else’s fault. Women actually have a more accurate judgment of risk in this particular way.

On the other hand, it’s called the male hubris, female humility effect. Men have more hubris. Women, in addition to having a lower hubris, also have higher levels of humility. Humility means that in the face of actual success, you’re less likely to attribute it to yourself and you’re less likely to take advantage of it.

Knowledge at Wharton: Right.

Mollick: So we thought: If entrepreneurship is based in part on hubris, maybe this male hubris, female humility effect actually shows us something about why women are less likely to do start-ups. And there are a lot of reasons. There’s the fact that women tend to have a lower preference for start-ups. There are issues that we can discuss later of ongoing misogyny and social barriers to female entrepreneurship.

But even given all of that, women still seem less likely [to launch statups], so we thought maybe this was the reason. Then, we also looked at crowdfunding, which is interesting because we can watch a lot of people fail, succeed and try at things in this very transparent way.

In this case, what we did was looked at people who had succeeded or failed the first time they tried to raise money on Kickstarter. The people who succeeded in raising money were those who had a goal and raised more than that goal. If they were trying to raise $10,000 to produce a new kind of coffee cup, they raised $10,000 or $11,000 or $15,000. We also looked at people who failed — who tried to raise $10,000 to make a new coffee cup, but they raised zero dollars or $5,000. What we figured was that everybody is justified in trying entrepreneurship the first time, because you don’t know how good you are.

But whether you fail or succeed the first time, you’ve now learned something about your own ability to become an entrepreneur. The people who failed now have new information that they may not be so great at this, and the more they fail, the more it’s an indicator they didn’t do a great job.

People who failed, but raised $9,000 out of the $10,000 — great. People who raised zero dollars out of $10,000 should be learning the lesson that they’re doing something wrong.

Knowledge at Wharton: When you think about crowdfunding and the ideas on crowdfunding sites, people are investing in the idea. They’re not really investing in the person, at least at the outset.

Mollick: Exactly. So what’s nice is gender shouldn’t have a role in this….

What we defined as hubris is trying to start a company, trying a second attempt at crowdfunding, when you failed by a lot the first time around. We defined humility as not trying again when you’ve had a big success. So, say I’ve raised $15,000 with a goal of $10,000: I should be more likely to try again; that tells me I’m very good at this. If I raised zero dollars out of $10,000, it tells me I’m bad at this.

“People who failed, but raised $9,000 out of the $10,000 — great. People who raised zero dollars out of $10,000 should be learning the lesson that they’re doing something wrong.”

Knowledge at Wharton: Time to try something else.

Mollick: Exactly. And we found evidence of this effect. As people failed by larger and larger amounts, disproportionately, women more than men were discouraged from trying again. Now that’s a very rational outcome, right? Because you’ve learned something.

Knowledge at Wharton: Right.

Mollick: Men were much more likely to try — but also, on the other end, they succeeded by larger and larger amounts. Women — again, increasingly disproportionately — were less likely to try again than men. They were less encouraged by success and more discouraged by failure. Being discouraged by failure is completely rational. It benefits women individually, because they’re not engaging in doomed ventures. But it hurts them as a group because it means that you don’t have enough women buying entrepreneurial “lottery tickets.” So fewer “lottery winners” are women, and you don’t have as many successful entrepreneurs who are women.

You also have fewer role models and less of a good basis to start things with. We found that — in our sample, at least — if women were as immodest and as unhumble as men, and as overconfident, there would have been 30%, roughly — about 28% — more female founding attempts in our sample. That was a huge number of people being discouraged by this psychological characteristic. It explained a lot of the gap in the founding rates between women and men in our sample.

Knowledge at Wharton: How many different crowdfunding concepts did you actually look at in the first place?

Mollick: We started with everybody who has ever tried to raise money, and we narrowed it down, if I remember correctly, to several thousand attempts. So it was a pretty good set of numbers.

Knowledge at Wharton: We were talking before we went on the air about research conducted by [Wharton’s] Laura Huang that looked at VC start-up funding rounds, which found that an attractive man would be able to get more money than a woman, but a woman would be able to get more money than an ugly man. It’s interesting how the personal dynamics play a huge role in terms of the actual funding, and the potential successes of these different ideas.

Mollick: Yes…. We’re still trying to work through these things, but it’s complicated.

If you look at the number of companies started by women in the United States, about 38% to 40% of all companies have female cofounders. Now, I’m going to ask you to play a guessing game. Overall, 40% of companies are VC funded, but of those, what percentage do you think have female cofounders?

Knowledge at Wharton: I’m going to say it’s probably quite a bit lower, correct?

Mollick: Yes — somewhere between 2% and 4%. Terrifyingly lower, right?

Knowledge at Wharton: Yes.

Mollick: I also have a set of research trying to understand what’s happening there and how to improve it — and I’m also looking at crowdfunding for that. One of the things to note is that women in general are less likely to raise money than men in almost every circumstance. We talked about some of professor Huang’s research; there is also other research looking at bank loans and several other categories.

[However,] what we found that was interesting is, with crowdfunding as opposed to any other form of fundraising, with the exact same project — controlling for all of the factors — women are 13% more likely to succeed than men.

Knowledge at Wharton: On Kickstarter?

Mollick: On Kickstarter. Jason Greenberg of NYU and I have been trying to figure this out, and we thought the reason might have to do with the percentage of people who are actually funding projects.

“Being discouraged by failure is completely rational. It benefits women individually, because they’re not engaging in doomed ventures. But it hurts them as a group.”

If we take a step back and think about why women don’t raise as much money as often as men, there are a bunch of reasons. We talked briefly about misogyny, and that’s been in the press recently.

[Also,] women in general tend to start companies that are different than the kind of companies that men start. They are more likely to start a company out of the home; they are more likely to be in area like retail, where venture capital and bank loans are not as common.

There are a bunch of reasons, but one key reason seems to be — and this is a problem that we see in discrimination everywhere — something we call “homophily.” It’s the principle that “birds of a feather flock together.” People like people like themselves. VCs tend to be mostly male; they have friend networks that are mostly male. As a result, you have a very strong network of men who talk to each other, and it’s very hard for a woman to get access to these people.

Knowledge at Wharton: Especially if it’s a network that’s been in place for 10 or 20 years.

Mollick: Exactly. It doesn’t matter how proactive and feminist you are as a guy. If the network you are part of is mostly men, you’re just not going to see as many women’s projects, you’re not going to hear in your network about as many successful women. You’re not going to be able to do due diligence as easily. This has been a problem in a lot of fields.

One of the things that you hear about as a solution to this is: What if we try to increase the number of women who are venture capitalists? Can we solve this problem, because then we are reversing that trend?

… [To look into this] we did several experiments. We took a Kickstarter project that was very successful, and we created two exact versions of that project.

The only difference between the two was the creator — in one case, it was created by Jessica Smith, in the other case, it was created by Michael Smith. Everything else stayed the same.

Jessica and Michael are the most common names for millennials, and Smith is the most common last name. We also used some data from Princeton to pick two equally attractive individuals, so they were controlled for attractiveness.… Then, in a lab setting we showed a bunch of people one of these two projects.

What we wanted to do was figure out whether the project being created by a man or a woman made a difference, so we actually gave [our subjects] money that they could choose to donate to the project, and we asked them where the project quality was better.

We found out that men didn’t care whether a project was created by a man or a woman. There was no significant impact. At least in this case, it didn’t seem to move the needle. With women, it turned out to be really interesting. Two-thirds of women actually thought the project created by the man was better than the project created by the woman.

[However], we took a bunch of measures, and we realized about one-third of the women in the sample were what we called “activists.” These were women who knew that women were underrepresented in technology. They felt that women suffered from discrimination in this field, and they thought it was important to try and fix that. They thought either the government should help or they should help or it was important to try and change this. Those women were much more likely to [fund] a project created by a woman.

So all of the success that we found — the reasons why women were doing better than men [on Kickstarter] — came from a small group of women who were helping to support other women in areas where there was the most disadvantage for them.

Thus, the solution to the problem of “how do you increase the representation of women” turned out not to be from having more women participate, but from [involving] these activists who are actually out to help.

Knowledge at Wharton: Given the fact that not as many women go back a second time to do a different project, in some respects, the success that a lot of women have in major corporations or in other companies — that falls in line with the idea that they are more often looking for a traditional, stable company, compared to going out on their own and really taking the gamble.

Mollick: Yes. We are seeing successful female entrepreneurship, and all the evidence and research seems to show that having a woman on the board of your start-up or as a co-founder increases your chance of success, so it’s not an ability problem.

“It doesn’t matter how proactive and feminist you are as a guy. If the network you’re part of is mostly men, you’re just not going to see as many women’s projects.”

Similarly, when I look at outcomes in crowdfunding, gender doesn’t matter in terms of whether or not you’re successful in ultimately producing your product, [launching] a company, or any other factor.

[But] if there are not enough women doing entrepreneurship, then there aren’t enough women as funders, there aren’t enough women as examples, and that ends up discouraging women and doesn’t solve the problem.

Knowledge at Wharton: You mentioned that in a lot of cases, [success] is a matter of other women supporting female entrepreneurs. Are you seeing increasing numbers of those cases of women supporting other female entrepreneurs?

Mollick: One of the nice changes that we’re seeing is that there are more and more female VCs, there are more and more efforts to try and help women succeed, and there are even crowdfunding networks built for women, like Plum Alley.

… I think the situation is changing, but I think we have to be aware of what’s causing it. The solution seems to be in supporting fellow members of your group — and that’s not just for women; it’s any underrepresented group.

Knowledge at Wharton: I wanted to go back to something you were talking about earlier — about the level of success that either men or women were having with these projects: the amount of success, whether they were able to meet their goal, or they came close to that mark, or if they got very little. How did that all factor in?

Mollick: Overall, women did better than men. [The question became:] Are they doing better than men because it’s rough out there for a woman, so only the most persistent women are succeeding? If so, maybe the pool of women is better than the pool of men.

So we surveyed a lot of people in the crowdfunding world, and we were able to measure all those human capital factors. Did you go to college? Do you have kids? How long have you worked in the industry? Even taking all that into account, we still saw the same effect. So crowdfunding seems unique in providing more opportunity. One of the reasons I’m interested in it is its ability to help democratize access to capital.

Because [the problem is] not just gender. The mean distance between a venture capitalist and a company to invest in is only 80 miles. So if you don’t live by a VC, if you haven’t graduated from a good school, if you don’t know the right set of people, it’s hard — not impossible, but it’s much harder — to get access to venture capital.

That’s why crowdfunding is interesting — because it spreads the net much wider and lets anyone raise money. It’s not just women. It’s also people who live outside of major cities; it’s people who are amateur inventors.

Knowledge at Wharton: In other words, not people who are already involved in that realm, which is obviously where a large percentage of the cash is coming from.

Mollick: Exactly. Otherwise, the money keeps coming from the same people and going to the same people, and it doesn’t ever break out of that world.

Knowledge at Wharton: The aspect of overconfidence, though, is an interesting one because a lot of people assume that being overconfident is a negative, yet in some of these cases, overconfidence was actually a positive.

Mollick: Well, overconfidence is usually a negative individually. The way you calculate returns is, you look at the chance of you winning something, and you multiply it by how much you win. That’s why a lottery ticket is never a good deal: You have a one in 100 million chance of winning a million dollars. Based on that, you know it’s costing you money to play the lottery.

Knowledge at Wharton: Especially if you are playing it week after week after week after week.

Mollick: Exactly. In the same way, with entrepreneurship, on average, you’re buying a lottery ticket. Now, if you have an advanced degree from a top school, the odds are much better. I’ve got research on Wharton entrepreneurs that shows they do quite well.

But, the expected return from entrepreneurship for most people is negative because most businesses close, and even for the businesses that succeed, only a small portion of those founders end up becoming super rich.

Less than 2% of companies in the United States receive venture capital and go on to do an IPO and all these other things. So overconfidence individually means that you’re not assessing the risks properly, because you think you’re better than everyone else based on no evidence. As a group, though, overconfidence gets people to try things, and it advances the state of technology, the state of society.

“The solution seems to be in supporting fellow members of your group — and that’s not just for women, it’s any underrepresented group.”

So the collective failure of entrepreneurs is bad for those entrepreneurs who are failing, but good for us as people who want to benefit from innovation and creativity.

Knowledge at Wharton: You’re talking about hubris and overconfidence. These human traits aren’t the sorts of things that are easy to alter. It’s not like making an adjustment to a business plan. Is there a thought process that the numbers might change over time, and that the research you did could lead to changes, or is this a pattern that probably will be maintained a long time to come?

Mollick: First, I should mention, just so people don’t think this is too trait-based: For a long time in entrepreneurship, we’ve been studying every aspect of personality to try and figure out whether you can give someone a test to find out why they’ll be successful as an entrepreneur.

For me, at least, the good answer is: You can’t. We have no way of predicting entrepreneurial success. We can predict the chance of you going into entrepreneurship, which is what overconfidence helps predict.

Those traits seem pretty fixed, but the good news for anyone out there who’s thinking about being an entrepreneur is, I can’t tell you in advance whether you’re going to succeed or fail based on your personality.

There is a sense that maybe entrepreneurs are more likely to want to have control. That might be an issue. There are some other findings that entrepreneurs who are willing to share the wealth do better, but there’s not really a clear personality characteristic that applies. So I agree, the trait stuff is somewhat locked in place. But I think you can change your own sense of confidence by looking around and taking a realistic view of your own chance of success.

Knowledge at Wharton: Sure.

Mollick: Also, go about staging your start-up in a way that you learn information as you go, so you don’t, for example, commit all of your money to launching a restaurant without having tested your ideas first. There’s a method out there called Lean Start-Up that gets around some of these issues. But the best advice I have for entrepreneurs is: Realize you’re probably overconfident.

That’s OK, but do what you can to learn about your own chances of success or failure before you commit all your money and time, and quit your job to do this.

Knowledge at Wharton: Obviously, if the idea is a sound, fundamental one, in the right industry, it really wouldn’t matter whether you’re a man or a woman.

Mollick: Yes. A good idea and a good team tend to win. The lesson, though, tends to be that women still have it tougher than men in getting access to resources.

So the issue is not in launching a start-up, not in being able to run something successfully, but that they have trouble getting access to resources in most areas of funding. And that’s something we have to still work on.

These are marginal impacts, but they discourage women from trying. I think it would benefit everybody, given all of our research, to have more people be successful entrepreneurs — both for the people who are making the money doing it, and for the society that benefits from the results.