2 ms·
There's a lot of discussion here about changing the incentive structure of academia to prevent this kind of fraud, but based on the discussion I've read so far
by ModernMech 5y ago
There's a lot of discussion here about changing the incentive structure of academia to prevent this kind of fraud, but based on the discussion I've read so far I don't think a lot of people here understand the exact incentives that are involved, so I'd like to try to explain how academics actually operate.
We've all head of "publish or perish", and there is a general understanding that you have to publish papers to get tenure, so that's where the incentive is. But it's a little more complicated than that.
When you are hired as a tenure-track assistant professor at a university, you're given what's called a "startup" package. Assistant professors are essentially startup founders where their company is a lab. Your job as an assistant professor is to use that money to purchase equipment and to recruit PhD students (aka employees) so that you can launch your research agenda. This money pays for:
- rent taken off the top (essentially paying for your lab space and other costs the University incurs)
- a portion of your own salary
- your students' tuition
- your students' stipends
- conference travel and fees
- lab equipment (computers, desks, chairs, machinery, scientific apparatuses). To note here, for every dollar you spend on equipment, the University takes a percentage of that. So if I buy a chair for $100, the University will charge for example $10. This is a sort of tax for being affiliated with the University and getting free use of their resources (internet, libraries, subscriptions, etc.).
- As well as other things.
Crucially, this money is not free. The expectation is that you will in the next 6 or so years use that money to jumpstart a successful lab operation. A "successful" lab is one that is able to procure significant funding from government and institutional sources. Your first goal is to procure enough funding to recoup the initial startup package the University laid out. That's your bare minimum. Ideally, you want to procure funding far in excess of this by the time you're up for tenure, to prove to the University that your lab will bring in more dollars. Because you've got to earn your keep.
Notice there hasn't been much discussion of papers so far. The initial startup package can be quite significant -- on the order of millions of dollars depending on the research agenda. As an assistant professor, you need to bring in millions of dollars from funding agencies like the NSF to make that back in multiples. But the NSF budget was only $8 billion dollars in 2020, and they need to spread that money around to new assistant professors across the entire country. Therefore, funding rates usually range from 10-20% for a proposal, which is quite low when your entire future is depending on it.
This is where "publish or perish" comes in. When you are up for tenure, yes they'll be looking at the quantity and quality of your publications. However, you could have 1000 high quality publications, but if you don't bring in significant grant money, you will not be getting tenure. The way you get that grant money is by writing a good research grant proposal, and the way you back up and bolster those proposals is through publications.
This is why there is so much fraud in publications. Not getting tenure at this stage means your research career is essentially over. You are as good as fired from your current role, and few Universities will take you on as a tenure track professor when you've already failed at your current institution. At this point you're in your mid to late 30s, and your best years (in terms of research ideas) are already behind you. It's time to enter industry or become a lecturer.
So how to fix this?
1. More research funding for government agencies. Higher funding rates means less incentive to commit fraud.
2. A feedback loop of profits from industry to universities. Right now when technological advancements happen at universities, those flow out to the general public, then to companies, who figure out how to monetize them. The profits generated by that monetization are captured entirely by private corporations, even though the foundational innovations on which they are based are funded by Universities and by extension the public. If some of this profit were diverted back into Universities (maybe through increased corporate taxes), then this would help.
3. Lessen the incentive for Universities to essentially become hedge funds with schools attached. My University is currently sitting on a billion dollars that is locked up in investments. Harvard's endowment is currently 40 billion dollars. They could fund the entire country's research agenda 5x over with that. Yet professors are paying their own salaries with public money? And the University gets a cut? This is insane.
4. Offer an alternative off-ramp for academics who don't get tenure. If not getting tenure won't end your career you've spent your entire life building, then the incentive to commit fraud would be lessened as well.
Anyway, tldr; there's not enough money to go around, so fraud exists.
- OminousWeapons 5y agoI agree with everything you said. At the risk of being pedantic: > A feedback loop of profits from industry to universities. Right now when technological advancements happen at universities, those flow out to the general public, then to companies, who figure out how to monetize them. The profits generated by that monetization are captured entirely by private corporations, even though the foundational innovations on which they are based are funded by Universities and by extension the public. If some of this profit were diverted back into Universities (maybe through increased corporate taxes), then this would help. This isn't ENTIRELY true. Universities "own" research IP and they do have tech transfer offices which at least theoretically help them monetize their largest breakthroughs via researcher startups. Many large, well established labs will also perform collaborative industry research, effectively acting as a CRO, which they can charge a lot of money for (and the Universities can subsequently take overhead from). The huge, field leading labs with well known PIs will have multiple R01s and various grants but the vast majority of their funding will come from industry.
- ModernMech 5y agoYes, this is true. But I was thinking more along the lines of fundamental research. For example, DARPA spent a lot of money investing in transistor research in the 50s, which begat microcontrollers, which begat companies like MS, Apple, and Google. Those companies together are worth multiple trillions of dollars, yet wouldn't exist but for the publicly funded sector research that preceded them. Harvard's $40 billion looks piddly in comparison. And what kind of research does MS/Google/Facebook fund in the University? They'd like us to figure out how to optimize ad placements on their search engines and social media sites. I've seen far too many promising researchers swallowed by that kind of work, only to quit after realizing the direct application of their research is simply to make those companies more profitable while they are paid $25k per year and eating ramen 3x a day. We see the same thing happening right now in robotics. DARPA spent a lot of money jumpstarting driverless cars with the DARPA grand challenge in 2005. The researchers I know who were involved with that have left academia and are in the private sector now because that's where all the money is. There's no money flowing back in the academic direction even though there's a direct line from what they were doing in 2005 to what Toyota, GM, Google, Tesla, etc. are doing today. I'm not saying they shouldn't make profits and be worth a lot of money. They provide a lot of value. But the outcome definitely seems lopsided to me.