Monday, July 18, 2016

Honesty, integrity, academia, industry

[Note added 7/22/2016 below in response to comments]

Implicit in my last post about reputation in science was one major assumption: that honesty and integrity are important in academia. The reason I left this implicit is because it seems so utterly obvious to us in academia, given that the truth is in many ways our only real currency. In industry, there are many other forms of currency, including (but not limited to) actual currency. And thus, while we value truth first and foremost in academia, I think that in some areas of industry, even those perhaps closely related to academia, the truth is just one of many factors to weigh in their final analysis. This leads to what I consider to be some fairly disturbing decision making.

It’s sort of funny: many very talented scientists I know have left academia because they feel like in industry, you’re doing something that is real and that really matters, instead of just publishing obscure papers that nobody reads. And in the end, it's true: if you buy an iPhone, it either works or doesn’t work, and it’s not really a debatable point most of the time. And I think most CEOs of very successful companies (that actually make real things that work) are people with a lot of integrity. Indeed, one of the main questions in the Theranos story is how it could have gotten so far with a product that clearly had a lot of issues that they didn’t admit to. Is Theranos the rare anomaly? Or are there a lot more Elizabeth Holmes’s out there, flying under the radar with a lower profile? Based on what I’ve heard, I’m guessing it’s the latter, and the very notion that industry cares about the bottom line of what works or doesn’t has a lot of holes in it.

Take the example of a small startup company looking for venture capital funding. Do the venture capitalists necessarily care about the truth of the product the company is selling or the integrity of the person selling it? To me, from academia, I thought this would seem to be of paramount importance. However, from what I’ve been hearing, turns out I was completely wrong. Take one case I’ve heard of where (to paraphrase) someone I know was asked by venture capitalists at some big firm or another to comment on someone they were considering funding. This person then related some serious integrity issues with this person to the venture capitalists. To which the venture people said something like “We hear what you’re saying. Thing is, I gotta say, a lot of people we look at make up their degrees and stuff like that. We just don’t really care.” A lot of people make up their degrees, and we just don’t really care. A number of other people I know have told me versions of the same thing: they call the venture capitalists (or the venture capitalists even call them), they raise their concerns, and the venture people just don’t want to hear it.

Let’s logic this out a bit. The question is why venture capitalists don’t care whether the people they fund are liars. Let’s take as a given that the venture capitalists are not idiots. One possible reason that they may not care is that it’s not worth their time to find out whether someone has faked their credentials. Well, given that the funding is often in the millions and it probably takes an underling half a day with Google and a telephone to verify someone’s credentials, I think that’s unlikely to be the issue (plus, it seems that even when lies are brought to their attention, they just don’t care). So now we are left with venture capitalists knowingly funding unscrupulous people. From here, there are a few possibilities. One is that someone could be a fraud personally but still build a successful business in the long term. Loathe as I am to admit it, this is entirely possible—I haven’t run a business, and as I pointed out in the last post, there are definitely people in science who are pretty widely acknowledged as doing shoddy work, and yet it doesn’t (always) seem to stick. Moreover, there was the former dean of college admissions at MIT, who appeared to be rather successful at her job until it came out that (you can’t make this stuff up) that she faked her college degrees. I do think, however, that the probability of a fraudulent person doing something real and meaningful in the world is probably considerably less than the infamous 1 out of 10 ratio of success to failure that venture people always bandy about, or at least considerably less than someone who's not a Faker McFakerpants. Plus, as the MIT example shows, there’s always the risk that someone finds out about it, and it leads to a high-profile debacle. Imagine if Elizabeth Holmes said that she actually graduated from Stanford (instead of admitting to dropping out (worn as a badge of honor?)). Would there be any chance she would have taken her scam this far without someone blowing the whistle? Overall, I think there’s a substantial long term risk in funding liars and cheats (duh?).

Another possibility, though, is that venture capitalists will fund people who are liars and cheats because they don’t care about building a viable long term business. All they care about is pumping the business up and selling it off to the next bidder. Perhaps the venture capitalists will invest in a charming con-artist because someone not, ahem, constrained by the details of reality might be a really good salesman. I don’t know, but the cynic in me says that this may be the answer more often than not. One might say, well, whatever, who cares if some Silicon Valley billionaires lose a couple million dollars. Problem is, implicit in this possibility is that somebody is losing out, most likely some other investors along the way. Just as bad, rewarding cheaters erodes everyone’s sense of trust in the system. This is particularly aggravating in cases when the company is couched in moral or ethical terms—and in situations where patient health is involved, everything suddenly becomes that much more serious still.

Overall, one eye-opening aspect of all this for me as an academic is that while we value integrity, skepticism and evidence very highly, business values things like “passion” more than we do. I don’t know that an imposition of academic values would have necessarily caught something like Theranos earlier on and all the other lesser known cases out there, but I would like to think that it would. Why are these values not universal, though? After all, our role in academia is that of evaluation, of setting a bar that employers value. In a way, our student’s aren’t really paying for an education per se—rather, they are paying for our evaluation, which is a credential that will get them a job; in a sense, it’s their future employers that are paying for the degree. Why doesn’t this work when someone fakes a degree? When someone fakes data?

Here’s a thought. One way to counter the strategy of funding fakers and frauds would be for us to make this information public. It would be very difficult, then, to pump up the value of the company with such a cloud hanging over it, and so I think this would be a very effective deterrent. The biggest problem with this plan is the law. Making such information public can lead to big defamation lawsuits directed at the university and perhaps the faculty personally, and I’ve heard of universities losing these lawsuits even if they have documented proof of the fraud. So naturally, universities generally advise faculty against any public declarations of this sort. I don’t know what to do about that. It seems that with the laws set up the way they are, this option is just not viable most of the time.

I think the only real hope is that venture capitalists eventually decide that integrity actually does matter for the bottom line. I certainly don't have any numbers on this, but I know of at least one venture capital firm that claims success rates of 4 in 10 by taking a long view and investing carefully in the success of the people and ventures they fund. I would assume that integrity would matter a lot in that process. And I really do believe that at the end of the day in industry, integrity and reality really do trump hype and salesmanship, just like in academia. I don’t know a lot of CEOs, but one of my heroes is Ron Cook, CEO of Biosearch Technologies, a great scientist, businessman, and a person of integrity. I think it’s not coincidental that Ron has a PhD. For real.

Update in response to comments, 7/22/2016:
Got a comment from Anonymous and Sri saying that this post is overblowing the issue and unfairly impugns the venture capital industry. I would agree that perhaps some elements of this post are a bit overblown, and I certainly have no idea what the extent of this particular issue (knowingly funding fakers) is. This situation probably doesn't come up in the majority of cases, and it may be relatively rare. All that said, I understand that my post is short on specifics and data when it comes to funding known fakers and looking the other way, but I think it will be impossible to get data on this for the very same reason: fear of defamation lawsuits. You just can't say anything specific without being targeted by a defamation suit that you will probably lose even if you have evidence of faking. So where are you going to get this data from?

And it's true that I personally don't have enough anecdotes to consider this data. But I can say that essentially every single person I've discussed this with tells me the same thing: even if you say something, the venture capitalists won't care. In at least some cases, they have specific personal examples.

Also, note that I am not directly calling out the integrity of the venture capitalists themselves, but rather just pointing out that personal integrity of who they fund is not necessarily as big a factor in their decision making as I would have thought. My point is not so much about the integrity of venture capitalists—I suspect they are just optimizing to their objective function, which is return on investment. I just think that it's shady at a societal level that the integrity of who they fund is apparently less important to them than we in academia would hope. Let me ask you this: in your department, would you hire someone on the faculty knowing that they had faked their degree? I'm guessing the answer is no, and for good reason. The question is why those same reasons don't matter when venture capital are deciding who to fund.