How pay works

Broadly speaking, graduate students are funded by a combination of funding from three sources:

Your advisor. Your advisor has a research budget, drawn from outside grants from the government or other companies. They can use this budget to pay your tuition and stipend (salary).
The university. When you're employed by the university to teach a course as a TA/GSI, or when the university provides you a fellowship, that money comes out of the university directly.
Other sources. For outside fellowships (e.g. NSF GRFP), this funding is provided by outside sources. Normally, you have to apply for this funding yourself.

As an EECS grad student, your tuition and fees are guaranteed to be covered (though the source of your funding depends on specific circumstances, e.g. if you're receiving certain fellowships like the NSF GRFP). As a grad student, you also receive an additional salary or stipend for your work. There are several different ways you can be employed by the university to do your research/teaching work, and the specific title you're given determines where you get your funding from.

One difficulty with writing about funding in general is that funding situations can vary pretty drastically. In particular, fellowships have wildly different terms.

GSRs, GSIs, step levels, and appointments

The most common situation is to be employed as a graduate student researcher (GSR). In this scenario, you are essentially employed by the university for that semester/period to do research. If you are employed as a GSR, your funding comes out of your advisor's research funds.

You could also be a graduate student instructor (GSI) or teaching assistant (TA), which means that you're employed for a semester to help teach a course. GSI salaries are paid by the university.

Berkeley is currently phasing out the term "GSI" for "TA." However, as we explain in the section "Changes in pay to GSIs/TAs," these terms aren't _entirely_ synonymous. The formal pay code for TAs actually pays us _lower_ than the pay code for GSIs, and departments across Berkeley are using this as an excuse to pay us less than we're supposed to be paid. See "Changes in pay to GSIs/TAs" below for more details on this.

You may also hear the acronym **academic student employee (ASE)** mentioned. ASEs include GSRs, GSIs, tutors (normally undergrads who help with classes), and readers (people hired to read/grade assignments).

GSRs and GSIs can be funded at different step levels, which are standardized graduate student pay amounts set by the University of California for the whole UC system.

We've taken the liberty of compiling historical data on GSR pay, which has been manually compiled from info on this page regarding GSR salary scales from the UC Santa Cruz website. Even though the info is from UCSC, the numbers hold across all UCs.

At the end of 2022, a large strike held by graduate workers across the UCs resulted in changes in this funding structure. We will go over this in the next section. In a world where graduate students were paid for full-time work the whole year, this is how much each step level would be making yearly in salary:

Step level yearly pay over time, as listed on the UC website

However, to be clear: this is not how much grad students at each step actually make.

As a grad student, your job is to do research work during the school year (i.e. fall and spring semesters). You are compensated for this work either through a GSR/GSI appointment, and/or through fellowship payments. However, students aren't paid for full-time work during the school year. During fall and spring semesters, you're expected to split your time roughly 50/50 between research work (paid) and being a student (taking classes etc., which is unpaid).

Thus, during the school year, you're employed at a part-time (i.e. ~20 hrs / wk) rate. Additionally, during the summer, you have the option of being employed full-time by the university to do research work. Historically, the expectation is that if students wanted, they could take a full-time position working for the university, and be paid full-time, at full-time rates.

Anyways, here are the table amounts again, but showing monthly rates, and accurately reflecting part-time and full-time monthly rates. This is more accurate to what folks across the university are experiencing (but not quite right, as we'll discuss below!)

Note that these amounts also don't include tuition or fee remission. As we mentioned above, the EECS department guarantees tuition and fee remission for all students, and generally speaking across all departments, appointed GSRs and GSIs receive tuition and fee coverage.

Appointment levels are the percent of your time you're expected to do research work during a period, and thus what percent of the full pay amount you will receive. Thus, you would expect that during the school year, you'd be appointed at 50%, and during the summer, you'd be appointed at 100%.

Notice I said "full-time" three times! This is foreshadowing.

Importantly, each department at each school sets the policy for what step level and appointment level its students should be funded at, and whether this should change based on student seniority.

So, what step level and appointment levels did EECS students get? Prior to 2023, EECS students operated under the following guidance from the department:

EECS funded all students at GSR Step 9.
By contract, departments are not allowed to reduce students' step levels over time, only increase them.
Since GSIs are paid less than GSRs, the department required faculty to cover the difference between GSI and GSR pay (GSI top-ups)
During the school year, EECS grad students are paid for part-time work (i.e. 20 hrs/week).
- EECS grad students yet to pass their qualifying exam were appointed at 48% during the school year. This is a totally arbitrary amount. EECS PhD students were not doing 2% less work than candidates. The department just wanted a way to pay pre-quals students less.
- EECS PhD candidates (i.e. those who have passed quals) were appointed at 50% during the school year.
All EECS students were appointed full-time at 100% during the summer.

One consequence of the appointment tomfoolery is that more junior EECS students were actually paid ~4% less compared to their more senior counterparts during the school year, since it's 2% out of a 50% correct appointment.

We can look at how monthly earnings during the summer and school year have changed from 2013 to 2021: