In August 2018, Sherry Moss published a commentary in Nature that described something most researchers already knew but rarely saw written down in a journal with that level of reach.
"Working conditions in academic labs encourage abusive supervision," she wrote. "It is time to improve monitoring of and penalties for abuse."
The piece opened with a specific case: a young woman whose PhD adviser followed her around the lab, shaming her in front of colleagues, yelling that she was incompetent. She wanted to minimize contact with him. She felt trapped. Leaving would mean losing nearly three years of work.
Moss's argument was structural, not personal. The problem was not simply that some supervisors were cruel. The problem was that the system was designed in ways that made cruelty low-risk for the perpetrator and catastrophically high-cost for the target. Power was concentrated in a single relationship with almost no external accountability. Leaving was not a realistic option. Reporting was not a realistic option. Endurance was the only strategy available.
She was right. And eight years later, it is worth asking honestly: what changed?
What institutional monitoring actually looks like in practice
Moss called for better monitoring and stronger penalties. These are reasonable recommendations. They are also recommendations that depend entirely on institutions having both the will and the incentive to implement them.
The evidence since 2018 suggests that will and incentive are unevenly distributed at best.
Universities have reputational interests that are not always aligned with the interests of the researchers inside them. A tenured professor with a strong grant record and a long publication list is a financial and reputational asset. A PhD student or postdoc who reports that professor for bullying is a liability. Not because institutions are uniformly malicious, but because the incentive structure consistently points in one direction.
When monitoring is left to the institution, the institution becomes both the investigator and the party with the most to lose from a damaging finding. Formal grievance processes, ombudsperson offices, and HR departments are not without value. But they are not, structurally, the same as independent external accountability. And for the researcher facing a Moss-described trapped situation, the distinction matters enormously.
Why the trapped researcher problem is actually an information problem
The young woman in Moss's opening example felt trapped because leaving would cost her three years of work. That is a real and serious constraint. But there is a constraint that comes even earlier, one that the monitoring and penalties framework does not address at all.
She chose that lab before she knew what it was like to work there.
The information that would have been most useful to her was not available at the moment she needed it. It existed, in the minds of the researchers who had already been through that lab, in the experiences of the people who had worked under that adviser before her. But it was not findable. It was not searchable. It was locked in informal whisper networks that only reached you if you happened to know the right people and ask the right questions at the right moment.
This is the information gap that no amount of institutional monitoring addresses, because monitoring by definition comes after the problem has already occurred. A PhD student who enters a toxic lab and survives it may eventually trigger a formal process that results in a penalty for the supervisor. That outcome does not help the researcher who is currently in the process of choosing that same lab right now.
The structural intervention Moss was calling for needs to work before the decision, not after the damage.
What bottom-up transparency actually looks like
The alternative to institutional monitoring is information transparency built from the researchers who have direct experience, not the institutions that have structural conflicts of interest.
RateYourPI (rateyourpi.com) is a free anonymous platform where PhD students and postdocs rate their supervisors across six categories: mentorship quality, work-life balance, publication ethics, funding stability, international student support, and overall experience. Reviews are verified through institutional affiliation. The email used for verification is deleted the moment a review goes live. There is no record connecting a reviewer's identity to what they wrote.
This is not a complaint box. It is a searchable, permanent record of what working under specific supervisors actually feels like, built by the people with firsthand experience, available to the people who are about to make the same decision.
The distinction from institutional monitoring matters at every level. Institutional monitoring is reactive: it responds to complaints after damage has occurred. A review platform is proactive: it provides information before the commitment is made. Institutional monitoring is controlled by the party with a conflict of interest. A review platform is controlled by the researchers themselves. Institutional monitoring protects the institution's reputation first. A review platform exists solely to serve the next researcher.
What Moss's trapped researcher would have needed
The young woman in the opening of Moss's 2018 Nature commentary did not need better institutional monitoring after she was already in that lab. She needed honest, anonymous information about that adviser before she signed the paperwork.
If the researchers who had already been through that lab had somewhere permanent and searchable to leave an honest account of their experience, verified enough to be credible and anonymous enough to be safe, she might have made a different decision. Or she might have made the same decision with open eyes, knowing what she was entering and preparing accordingly.
Either outcome is better than what the trapped researcher actually faced: years into a situation she could not have predicted and could not safely leave.
Eight years is enough time to know what institutional solutions do and don't fix
Moss was right in 2018 that the structure of academic supervision enables abuse. She was right that monitoring and accountability mechanisms needed to change. Eight years of evidence suggests that institutional solutions have improved things at the margins, but have not changed the fundamental dynamic she described.
The young woman following a researcher around a lab, yelling about incompetence in front of colleagues, still happens. The researcher who feels trapped still cannot easily leave. The information that would allow the next person to avoid the same situation still does not flow through any formal channel.
What has changed is that the infrastructure to fix the information problem now exists. The monitoring Moss called for does not have to come from the institution. It can come from every researcher who has direct experience, contributing anonymously and permanently to a record that the next person can actually find before they commit.
That is not a replacement for institutional accountability. It is the piece that institutional accountability, by its nature, cannot provide.
Sherry Moss's commentary Research is set up for bullies to thrive was published in Nature in August 2018. If you have worked under a supervisor worth rating, your anonymous review at rateyourpi.com is exactly the kind of information that changes what the next researcher knows before they decide. 30 seconds. No login. No trace.