Should I make my research students' pay contingent on completion of tasks (such as reading/summarizing papers)?

Question

I received a fund that I can use explicitly for paying graduate (master's and doctoral) students for research. These students already receive a stipend from the university, but most advisors who can afford it pay them a supplement. My colleagues just pay a fixed amount per month. However, I am inexperienced in advising students, and I fear I might end up wasting the money on students who do not do their job.

So I thought of making the payment contingent on some research-related task. I do not want to make it contingent on actual research results, since the "flow" of results (particularly in my field - theoretical computer science) is not constant, and the students need a reliable source of income. Instead, I thought of paying the students for reading and summarizing papers and books related to their research topic (up to the maximum amount I can pay per month). There are several reasons:

• Reading is an important part of research.
• Most fresh graduate students do not read enough - they prefer to program. They need an incentive to start reading.
• Reading is a relatively "stable" activity: by putting a sufficient amount of effort, the student can guarantee a fixed income for themselves.
• Summarizing research papers will help the students once they start writing their own papers.
• The summaries can also be useful for me during teaching, as supplementary material for the students.

Is it reasonable to pay research students based on reading & summarizing papers, or for other activities?

Clarifications:

• The basic income (before the supplement) may or may not be enough to live on. For students who come straight for undergrad, many are young and still live with their parents. During undergraduate studies they had to pay tuition, so the very fact that they now study for free (even before receiving any money) is a substantial upgrade for them. Other students are older and have children; for them, the basic income paid by the university may be insufficient.

Answer 1

Do not do this. At best, you will appear merely stingy; at worst, perhaps financially abusive. (I also suspect your university would not even allow it, but I think that's not as important because it should not be done anyways)

I strongly, strongly advise you to reconsider how you think about students. It seems you are thinking about them like employees you want to squeeze some work out of, with some sort of benefit that you deserve for providing funds. This thinking is wrong, and leads to many of the student horror stories here about abusive advisors.

A student should never be under threat that their financial security is subject to the whims of their supervisor on a regular basis. Any process that leads to a student getting paid less than their full salary* should be a structured process with warning and opportunity for improvement. You write that "students need a reliable source of income" - exactly, so don't make it unreliable.

Students in research are not mere employees, they are future researchers for whom you are taking on a responsibility as an advisor. You pay students so that they can focus on their research and education. You pay students who you think have good potential as researchers that are worthy of your time investment. You are not paying them to do a job, you are paying them to learn.

If you think you need to add incentive for students to focus on things like reading, do that incentivizing outside of the regular compensation structure. One option is to have regular meetings where students present papers or literature reviews of a subset of papers. Students are typically quite motivated in that setting to look good in front of their superiors and peers and they will put in the effort. You could also use a bit of peer pressure to offer some sort of group reward, like buying lunch for the whole group when a student wants to schedule a presentation over lunch to summarize what they have read (harder to do now during pandemic times, but hopefully that won't be permanent).

I'm glad you've asked about this here before implementing it, because I cannot stress enough how bad this would be.

*Exceptions: for someone who is a part-time employee, like an undergraduate student, there is nothing wrong with hourly compensation, but it should not be based on production but rather on time worked. You can decide at the end of a semester if it's worth taking that student for another semester. There could be graduate students in special circumstances that work and study part-time; they should be paid a salary relative to their part-time level of commitment.

Answer 2

Not a good idea, and here's a take from a behavioral point of view, taken from this article

In Haifa, Israel, there were a bunch of day care centers, about a dozen of them, and the kids came in the morning and the parents picked them up in the evening. And as is the case, some parents came late. And they decided that they were going to impose a fine on the parents coming late. Well, happily for economic science, a couple of behavioral economists knew about it, and so they said, now, wait, wait, wait. Let's just do it on half of the day care centers and not do it on the other half so we have a nice experiment. So, sure enough, one day, the parents came, dropped off their kids and there was a notice. It said, starting tomorrow, anyone who's picking up their kid more than 10 minutes late will be fined 10 Israeli shekels. And then they recorded. They had been recording what was happening in the week before, and then they recorded how many people came late in the day care centers which had the fine and those where there was no fine. It was amazing what happened. In the places where there was no fine, nothing happened. It continued. There was actually a small number. In the ones in which the fine was imposed on the parents coming late, the amount of lateness doubled - doubled.

Now what - how can you possibly explain that? The fine was supposed to get them to come on time to pick up their kids. Now, if you think about it, there are lots of possible interpretations of what happened. But what you just said, Shanker, about framing seems to be the most likely explanation, which is the parents framed coming late or coming early to pick up their kids as essentially a moral question. I mean, perhaps not high morals, but you ought to pick up your kid on time because your kid might be anxious, because the teachers maybe want to go home and be with their kids or there's something like that. OK. Sometimes there is extra traffic, and you're late. But it was a moral question. As soon as you put a price on it, then it's just like a commodity. It's a shirt or a beer. Step right up. You want to get some lateness, here's where you can get it. It'll only cost 10 Israeli shekels. So I think they turned this thing from an ethical problem into more or less a self-interest problem. And, apparently, 10 Israeli shekels wasn't a big enough fine to really cause them to do anything differently.

The core idea is that, if something that is supposed to be done because it is the right/good thing to do, is incentivized with money, then it starts to become a commodity, and loses its original purpose. Feeling like you've got enough money this month and don't like to read? Just don't read and don't get paid, nothing wrong.

Motivating students to read should come from the actual value of doing it, like what you have mentioned in your post: that reading papers and summarizing them is useful. So, I believe it's better to have the funding be paid consistently, and not prorated based on the number of tasks they did.

Answer 3

Beware of setting targets. If you set a target, and an incentive for reaching it, then meeting that target becomes the most important thing.

Graduate students should be doing research, writing papers and writing their thesis, in addition to the background reading.

But if you pay them to read papers and write summaries, then that becomes the most important thing. Other activities can, and will, be left until their target for reading and summaries has been met.

Answer 4

In theory it sounds fine, but I think you will encounter a number of practical problems.

By paying for summaries you are incentivizing students to skim the paper and whip up a summary with minimal effort. I'll go ahead and assume your students are not so unethical as to hand you fake summaries, but I think it's evident that you will need to define quality guidelines for the summaries to "count". But if you do so, you will then have to verify the summaries to make sure they are of sufficient quality. It's not impossible and I have successfully done somewhat similar things, but it requires setting up a sophisticated system beyond just paying per summary. All that just to get the student to read... The student came to you because they have confidence in your mentorship. Can't you just talk to them and ask them to please read the damn things by next monday?

Moreover, I am a bit skeptical that money is a strong motivator for your students. If they cared about optimizing income, they would have gotten much higher paying industry jobs (especially in your field). Even if money has significance to them in a career sense, like any young student they are probably focusing on getting their career started so they can do well later rather than chasing pennies now. Often the feedback from advisor and the thought that their future career prospects are improving is a better motivator. Not like those are excluded by you paying them extra for writing papers, but you shouldn't get distracted from a big thing by focusing on a minor thing. Concerns about micromanaging aside, if you believe reading and summarizing papers is important you can simply require students to do it, rather than giving an option to skip it and getting paid slightly less. Maybe you could ask for a 5 or 10 minute presentation of the paper at every meeting? That would also help them practice presentation skills and disseminate the content to other members of your group.

Beyond this, you also risk inviting a disproportionately negative response from both your colleagues and the students. See for example the other answer - people in academia do not like this kind of incentive based scheme, they prefer to be paid anyway and have freedom to do their work as they see fit. Even if you had a system that was somewhat effective, you could have so much resistance that it negates any benefits. On the same note, because in academia you are rarely incentivized in this way (even publish or perish tends to take the form of gating at the application stage or annual quotas, not pay-per-work) so in a sense you are wasting the student's time by teaching them a way of working that is unlikely to translate into their future career.

As a tangent, learning to read and summarize papers quickly is an important skill for researchers. But if you want to teach this the "hard way", it may be more practical to simply limit the time they have to do it ("here's a paper, write me a summary by the end of the day"). Their first few attempts will probably not go well, but once they get a chance to experience and reflect on why they fail to do it quickly, you can start teaching them how to overcome those problems. While it may sound harsh, nothing stops you from being as nice about it as you want, such as asking the student to schedule some time for this in advance ("can you find a day this week where you have 1-2 hours for our paper skimming exercise?"), being encouraging, telling them that it's okay to not be able to finish in time and this is just a learning exercise, etc.

Answer 5

The other answers cover a lot of different perspectives which are quite important. I am not in academia any more but I have always closely watched the related issues.

Firstly, try to understand, why the funding agency is funding you? What are their vision and specific purpose of this fund? Is it to help you as a new early-career researcher or is it to help you help the students? It seems like it is for the later.

When a fund is being utilized all stakeholders would like to see a tangible result at the end. Students must have learnt something, you must have achieved a research objective and the funding agency must have seen a positive outcome towards realizing their vision.

With this broad guideline, if it suits you, you could try to make decisions. The other answers seem to be from academicians and are at a more micro level.

Answer 6

I'll provide a meta-answer for your question. I believe you are facing this dilemma due to a combination of several factors, with the most important being the following:

1. Your academic institute at Ariel is in a legal twilight zone - the Israel-occupied Palestinian west bank; and is itself not a regular part of the Israeli academic system (e.g. the Israeli Council for Higher Education), so employment conditions overall have some aspect an "anything goes in the wild west". (Even the question of whether it's a college or a university is a subject of political debate.)

2. Junior researchers are (AFAIK) not unionized at Ariel college/university, nor in Israeli universities proper. The academic staff unions threw them out in the 1970s, and the universities degraded them from full-time academic employees to being only employed in teaching and receiving soft money in all sorts of ways for research. A description of this process at the Technion can be found in my chapter in the book on "Wrongful employment" (העסקה פוגענית); available online ; see chapter 16. Over The current unionization picture is that junior academics are mostly unionized as teachers, but those unions avert their eyes when it comes to research work.

   Had your employees - and they are employees, even if they're also students - been unionized, you would not have had the prerogative of choosing how to employ them: It would have been in basically the same format and under the same conditions as everywhere else.

3. The Israeli labor courts - who are infamous in their bias in favor of employers - have a checkered and inconsistent history of ruling about employer-employee status of people who perform useful work in relation of, as part of, or during their studies, or as a condition for being recognized with some academic degree. If you try to extract a consistent line of reasoning, all of them should be employees; but the courts often don't rule that way, contradicting earlier precedent. This also means they have never tried to shape a consistent doctrine for the status and rights of "not-quite-employee student-workers" - despite supposedly recognizing the existence of this category. So you are without a legal yardstick even ignoring the question of occupied territories vs Israel per se.

4. Perhaps this goes without saying, but - lack of oversight w.r.t. employment practices and their ethics, by the funding bodies.

As for what you should do: You should pay junior researchers as you yourself are paid - by hour (or an FTE-fragment) spent doing research-related activity. That includes reading, and summarizing papers (assuming you find that to be a useful part of the research process), and writing, and probably just sitting at the office scratching one's head thinking of abstract problems. Of course, they're not quite like you in that they're not independent researchers and are supposed to be guided by you, including in the allocation of their work time to some extent - but you (ethically) cannot condition their payment on their research achievements. An exception is prize/bonus money, over the basic guaranteed wage - which you yourself may be competing for as a senior academic staff member; although, frankly, I'm not a big fan of that either.

Answer 7

I'm going to detract from the other walls of text here and suggest that it may be worthwhile to use financial incentives to encourage participation (though by grad school a student should be self-motivated) - provided you establish a minimum baseline and use the incentives as bonus.

Just be careful as this is arguably a power dynamic which may be abused. It depends critically on your good faith.

I wasn't paid during grad school; students aren't entitled to this money, unless things are different where you are, though it may make studying easier for some. If you're having problems with unmotivated students, as you allude, then this may be a viable option, if unorthodox.

Answer 8

Beside the legal and psychological perspectives, this is contrary to how an efficient labor market works. (Although in many circumstances, in reality you do get labor markets which allow for this kind of contract due to the concentration of power) You seem to want freelancers, not employees - or worse, to have somebody enter a contract which combines all the downsides of being an employee with all the downsides of being a freelancer, while you have your cake and eat it too.

In a employer-employee situation, it is the employer who bears the financial risks. The employee gets a lower payment than a freelancer, and has a secure income. They are not involved in the enterprise's losses (or success). They are being paid for turning up and doing work, no matter what the result.

In a freelancer-client situation, the freelancer functions as a one-person enterprise, and thus bears the risks of business. You contract with an enterpreneur not for a volume of their worktime, but for achieving a result, and pay for it. Accordingly, it is up to the enterpreneur to manage all financial and complexity risks connected to the project. In some countries, this is even embedded in the legal system (Arbeitsvertrag vs. Werkvertrag in Germany).

Obviously, nobody wants to bear a risk. Just the possibility of making a win if things turn out easier than expected is not sufficient to weigh up for the potential losses. It is you who wants the result, and it is your risk that somebody sitting with a paper may not produce the results you wanted. If you want somebody else to bear that risk, you have to pay for it - in practice, this functions through freelancer rates being much higher than employees'. (I am handwaving here a bit about real life complexity - in today's regulated labor markets, there are of course other factors behind freelancers' higher rates, such as health insurance rates. But you'll note they also tend to be connected to bearing a risk). So, if you want to not bear the risk of "when my instructions were followed, the result turned out to be not useful for me, or was not ready in the expected time", you have to 1) find the kind of person who is willing to bear that risk (and so appears on the labor market as a freelancer, not as an employee) and 2) pay market rates for it.

There is a second risk you want to guard against - a lazy student. The economic way to state it is that in a situation of incomplete information, Adam Smith's theory of perfectly efficient markets doesn't hold, and you are in a classic principal-agent theory. This is not something solvable by forcing result-only contracts on people who apply as employees, first because it has the side effect of transferring other, very large risks wihtout a payment, and second because of all the non-economic reasons covered in the other answers. The way it works is by trust. You first hire a student whom your instincts qualify as trustworthy, then you wait for the execution of tasks, and if you realize that the student has been taking your money without spending any time working, you fire them. If you want the theoretical perspective, this is also borne out by game theory - it roughly corresponds to the tit-for-tat strategy in multi-round prisoner's dilemma (which is an optimal strategy). Without the theory, this is how millions of employers and employees solve it every day.