The tenure-track system in India: A systems researcher's perspective

Academic jobs come with a lot of freedom. Typically, you teach one course per semester, and the rest of the time is supposed to be spent on research. But in a non-tenured system, it's easy to become complacent over time. Maybe you hit a rough patch in research, or other factors come into play. You can start skipping office hours, keep teaching the same course for years, be lenient in grading -- basically, do just enough to keep your job without raising any alarms.

The tenure system tries to prevent this by evaluating faculty performance yearly for the first six years. The evaluation is based on three factors:

If you perform decently across all these areas for six years, you get tenure. After that, though, you're pretty much free to enter the same complacency mode.

Why the tenure system doesn't work well for systems research

On the surface, tenure sounds like a good way to keep faculty accountable. But implementing a fair tenure system is tricky, especially in systems research. One major issue is the expectation to publish every year.

In systems research, solving a meaningful problem takes time. Building a solid research prototype and submitting a paper usually takes two to three years -- assuming everything goes smoothly. If results don't pan out as expected, you might need to switch directions or spend even more time making your work publishable.

Some research groups manage to publish multiple papers per year. But there's a catch -- they get applications from top students worldwide, often from well-established master's or pre-doc programs. Some of these students already have good publications before even starting their PhD. These groups also have access to significant funding and don't face restrictions on the number of PhD students they can take in.

The PhD student problem in India

Indian universities, on the other hand, struggle to attract strong PhD candidates. Many times, if you're looking for a PhD student, you just won't find anyone suitable. And if you do manage to recruit one, they'll take around two years just to get trained enough to work on a problem.

Systems research requires a lot of effort, and there's always a high dropout rate. About 50% of students who aren't motivated or hardworking will leave the PhD program within the first three years, meaning you'll have to start all over again.

Then there's the issue of student limits. Indian institutes usually restrict the number of PhD students a faculty member can supervise -- often just one or two. If you want to take in more, you'll need external grants, which aren't easy to secure. Funding for government institutes is a bureaucratic mess, and in many cases, success depends more on who you know than on the quality of your proposal. So, the real challenge in Indian academia isn't just publishing -- it's finding good students and getting funding.

How some people bypass the system

Given these constraints, publishing one paper a year is tough. But there are workarounds. Some researchers get around this by being "guest authors" on papers from well-connected research groups. If you attend their meetings and know them well, they might include you as a co-author -- even if your contribution is minimal.

Now, not all collaborations are like this. Real collaborations happen when solving complex problems that require expertise from different areas. But in a tenure-track system that mainly looks at publication counts, it's hard to differentiate between meaningful collaborations and guest authorship.

The problem with teaching evaluations

For teaching evaluations, most institutes rely on student ratings collected at the end of a course. Some tenure models blindly use the average rating as a measure of teaching effectiveness. But here's the issue: if students don't regularly attend lectures, how reliable is their feedback?

Attendance is a problem in many courses these days. In a class of 300 students, maybe only 50 show up regularly. One way to increase attendance is to introduce in-class exercises, but this often backfires. The unmotivated students start complaining and making so much noise that even Dumbledore wouldn't be able to silence them. So, most faculty just let things be. Administrators aren't keen on enforcing mandatory attendance either. They want to "encourage students to learn on their own," even when evidence shows that attending lectures improves grades. And they won't change their stance unless the foreign universities they're copying from also start facing the same problem and decide to fix it.

Given all this, it's not surprising that faculty sometimes manipulate their ratings. If tenure depends on student feedback, it's easier to secure good ratings by being lenient rather than actually improving teaching. Also, in a country with millions of students, it's frustrating that we still struggle to find 300 who are genuinely interested in computer science.

Service: The least controversial part of tenure evaluation

The "service" component of tenure evaluation is mostly about administrative duties, serving on program committees, or giving keynotes. It's fairly objective compared to teaching and research, and while it matters, it's not usually a deciding factor in tenure decisions.

The fundamental issue: A number-driven tenure process

For systems researchers, the biggest challenge in securing tenure is that evaluation committees mainly focus on publication counts. Some institutes won't even let you apply for tenure unless you have six core-A or A* publications during your tenure period.

This happens because deans and administrators, unfamiliar with specific research areas, rely on simple metrics to make decisions. Counting publications in ranked venues is an easy way to appear "objective," even though it doesn't reflect the actual quality or impact of the research.

The need for a smarter tenure system in India

Despite its flaws, the tenure-track model does solve the problem of academic complacency. The issue in India is that universities have imported the model without adapting it to local challenges. The system values quantity over meaningful research, forcing faculty to chase numbers rather than focus on solving real problems.

A better approach would be to evaluate research based on long-term impact rather than publication counts. But for that to happen, leadership positions need to be filled by people with a broader research vision -- people who understand the realities of academia and are willing to push for meaningful reforms. Until then, faculty will continue to navigate a system that rewards metrics over actual contributions.