On the inherent injustice of admissions
Yesterday was the day I told more than 600 students no. No, you may not learn with us. No, you may not join our community. No, all of that hard work you did to try to gain access to our major was not enough. No, you made it to the University of Washington, but you can’t join the Information School. As the chair of the admissions committee, it feels terrible, because deep down, I know that all of these rejections are unjust: every single student at the University of Washington that wants to learn about information from us, and perhaps even beyond the university, deserves to learn from our faculty.
Of course, as we all know, “everyone” does not scale. I can teach one on one pretty well; I do it all the time with my doctoral students. But my lab has limits: when I go past 6 or 7 of them, things break down. But if I make some compromises, such as bringing together a large group of students together (otherwise known as a class), I can scale that teaching to perhaps 30–40 students, knowing each of their names, their dreams, and how I can help them. These efforts to scale are how we take an inherently fixed resource—my time—and try to distribute it to as many as possible.
But there are limits. At some scale, my ability to develop a meaningful connection with a student disappears. When I teach a class of 100, 150, 200 (or at the extreme, a massively open online course with 20,000), I am no longer a person to those students that knows them, cares for them, or guides them. I am just a distant voice in their head, like any other YouTuber, someone to be admired perhaps, but not someone to have mutual exchange with. This is the cost of scaling: a loss of connection, of community, and of commitment.
Since I’ve been in charge of admissions to our major, and learned from the experiences of so many other faculty who have led admissions processes, I’ve come to the conclusion that there are only two ways to resolve this tension. The first is to increase the amount of attention. This is how public schools in the United States work: when a neighborhood grows beyond its capacity to teach in a relational way, it builds more schools and hires more teachers. It recognizes that youth need meaningful, personal attention to some degree to thrive, and that teachers can really only provide that meaningful attention at the scale of perhaps 25 students. And some would argue (or present empirical evidence) that 25 is even too many. When classrooms scale beyond that, districts go back to the public and say, “We need more money to teach your children well.” And at least in communities that value education, they often get it.
Colleges and universities resolve this tension differently, usually by limiting demand for attention from faculty rather than increasing its abundance. To do this, we erect admissions barriers for deciding which lucky few will get our attention. We have admissions to our institutions, we sometimes have admissions to specific majors, and and we have enrollment caps on specific classes. In the absence of more attention to distribute, these admissions mechanisms focus the attention we have, at the expense of inclusion.
But after spending years trying to devise admissions mechanisms that are more inclusive, I’ve come to the conclusion that they are inherently unjust, and cannot be fixed. This is because they start from the premise that there is some fair way to decide which students deserve our attention. This premise has grown into the myth of meritocracy—that some students are more deserving because of their skills, their wit, and their experiences. And of course, there is some truth to this myth: students do vary along these dimensions. But part of this that is myth is that all of these differences are a byproduct of the students’ effort alone. Indeed, some students work harder than others, but whether that work translates into confidence and ability is greatly determined by what resources a student has to support them, whether mentoring from a parent, a helpful tutor, the luck of having an inspiring teacher, or simply having a functional school system. Thus, when students pursue college, they come with a prior knowledge that is a product partly of hard work, but partly the opportunities granted to them by their position in society, and more importantly, the position of their their guardians and their community.
The other part of this myth is that effort, and the ability that comes with it, are enough to justify attention. This moves us from empirical facts about equity to more philosophical claims about values. Public education in the U.S. and other countries, for example, starts from the value that every child, regardless of who they are, who their guardians are, or what resources they have, should have access to education. Our values in the U.S. on this front are weaker—we don’t guarantee a high-quality, well-resourced education as in most European countries—but they do start from the same core value. Most colleges and universities, in contrast, start from a different value of merit, arguing that those who deserve an education are ones who have earned it, whatever that means. Our only exception to this is the patchwork of 2-year colleges, which attempt to serve everyone. Thus, our decisions about who “deserves” faculty attention are value-based ones, not objective ones.
Starting from a value of merit leads to the morally fraught mechanisms we all know so well. For example, how does one judge merit? Most common is to attempt to measure demonstrated knowledge in the form of grades. We tell students: participate in a series of educational games, then complete some tasks and we will judge them on a scale. If you score high on that scale, you will have a higher chance of of getting our attention. This prizes the ability to play those games well, and ignores the reality that playing those games well requires more than just effort, but also health, wellness, time, money, and support. Therefore, a low grade might mean that someone lacks knowledge, but it might just as easily mean that someone was sick, depressed, overworked, poor, or isolated. And a high grade might mean that someone worked hard and knows a lot, but it might just as easily mean that someone has a rich community of support, a lucky streak of health, a stable community of supportive parents, family, and friends, or that they cheated the game. Trying to measure merit partly means valuing effort, but it also means valuing privilege, class, wealth, and luck.
If colleges and universities must work with limited attention, there are alternative values. For example, one could strictly value equal opportunity over effort. This might come in the form of a lottery, in which everyone that wants attention requests it and has an equal chance of getting it. At a conceptual level, this seems fair, but at a visceral level, students, parents, and society hate it. By ignoring effort, it sends a strong signal that no one person is more deserving of attention, no matter how hard they work or how clever they are.
In designing our Informatics admission process, I’ve tried to strike a balance between values of effort and equal opportunity. Our major is a “capacity constrained” major, which means that even though students have already been admitted to the University of Washington, if they want to complete our undergraduate degree, they must also apply to our program and be admitted. To strike a balance between effort and equal opportunity, I started from the premise that everyone is deserving of our attention, but not everyone will benefit equally from our attention, because we are not equally capable of supporting each students’ interests and aspirations. And so we ask students to carefully articulate their interests, their aspirations, but also their experiences with information, and the evidence of their learning that they believe positions them to pursue those interests and aspirations. We read their statements, carefully looking for evidence that 1) they have learned what they need to pursue their aspirations, 2) they understand how our program might help them with those aspirations, and 3) that our expertise and resources can actually help them achieve those aspirations. (Of course, since they have to write this clearly for us to understand, we are also implicitly judging their ability to write clearly). These criteria result in judgement of how well we will be able to serve each student, given their goals and our expertise. We rank students by that judgement, and admit as many as we can feasibly serve.
In principle, this works okay. I’m confident that we admit students that we can serve well, and that we don’t admit students who need different teachers than us. But in practice, it is carnage. We regularly have more than 900 students apply for about ~250 “slots” worth of attention. And so we say no to 75% of students, even though all of them would learn and thrive to some degree. And worse yet, of those 75%, the majority really are students who we could not just serve, but serve well.
So is this process just? No. All it does is shift attention away from grades to the alignment between students’ dreams and our ability to help students pursue them. No, justice would be growing our school’s faculty, and therefore its attentional capacity, to effectively serve all 900 students. Justice would be letting everyone into the University of Washington that wants to be here, and serving the thousands of students that want to learn about information. Justice would be me being free to say to every student: “You are worthy and we are here to teach you. Welcome.” But instead, I must say, “You are worthy, but we chose someone else to teach, because we thought we could help them better.” That is not justice. That is just a slightly more fair statement than, “You didn’t play this educational game well enough, try again.” At least with the educational game, there is a path to increasing their odds. In our system, there’s not a clear path forward when we tell a student, “We didn’t think we could help you as much as another student.”
What is the way out of this? In my view, there are essentially three options. One is to simply divide our attentional pie into smaller slices. Grow class sizes, teach all students that want to learn, but abandon the notion that we can develop relationships with every student as an individual. They become a row in a database and we become a talking head. To a degree, we’ve already done this; I do not know all of our majors as people, let alone all of their names. The scale we’ve achieved means that those relationships are distributed across our faculty and that there is no longer one community, but many. And this is what most other universities do: they let a student declare any major they want, but they make no guarantees about a student ever truly connecting with faculty in a mutual way. Many students would make this trade, but for the wrong reasons: they would take it because they believe that the value of education is content, and it would provide them access to content. But students would be wrong: the true value of an education is relationships, and those do not scale. This just shifts the injustice of exclusion to the injustice of an industrialized model of education, in which we deposit knowledge into students’ minds, one by one. That is not an education.
A second option is to grow. More faculty, more attention, more students. We’re doing our best to do this with limited support and resources, but we keep bumping into barriers. We’re out of office space for faculty and staff. We keep encountering the zero sum game of higher education budgets where every student that we admit to our major is a student lost to another, which just moves money around from one department to another. If we grew enough to serve everyone, it would mean the financial devastation of other disciplines that some students fall back to, like Geography, Math, Statistics, Design, Physics, and others. This just trades student competition over faculty attention for departmental competition over student attention, all while eroding the necessary diversity of academic expertise.
The third option, and in my opinion the only just one, is to grow the financial pie to meet demand. Guarantee that all departments, independent of their demand, have enough funding to meet demand and to maintain a thriving presence of their discipline. Because every department would need baseline funding to exist, that would mean that the university budget might have to increase when certain disciplines experience a rise in popularity. In fact, this is roughly what happened in Computer Science at our university recently: rather than taking money from humanities and giving it to CS, the state (and more so, philanthropists) gave CS more money to grow. This is what should be happening for all high-demand disciplines. But the public isn’t demanding access to our Information School, so our state legislature isn’t providing funding to meet demand, and so we are left to find our own way to grow.
And I am left with almost 600 students, writing to me emails, messaging me on Reddit, talking to me on Zoom: why? Why wasn’t I good enough? What do I need to do? One of them, last night, wrote on a Discord server that he was going to climb to the roof of his apartment and jump. He shared a photo of him on the roof. Some of our majors, caring and desperate to intervene, did their best to talk him down, to connect him with our university’s crisis line. And I sat in my room, trying to find the students’ location, their phone number, and to support the students in contacting 911 for a wellness visit, knowing that the inherent injustice of the process I oversee, and the rejection email he received with my name on it, was the emotional trigger for desparation. I wanted to tell this student: There is no why. You are good enough. Keep searching for another path, and you will find it.
He eventually went to bed. And the students found his roommate, and tried to convey the gravity of the situation, sharing with him the resources he might need to prevent this student from taking his life. And I went to bed, trying to remind myself that while I am in charge of the process, and I’m not responsible for its inherent in injustice. That responsibility lies with us all in The Information School, at the University of Washington, and in Washington state, who decided that while we will help every student find their dreams, we will not help every student pursue them.
If you are a student that we did not admit: you are good enough, and I’m sorry we could not serve you. Informatics is just one path of many. Some of those paths will take you to where you want to go, other paths will take you to places you haven’t imagined. Life is a full of closed doors, but it’s more full of open ways. Keep searching for those open doors, and keep searching for people who will help you find them.
And if you feel up to it, organize and demand more resources from a state that fundamentally believes in the importance and value of information, information technology, and information experts to shape a more just society, but doesn’t fund it. With the financial devastation this pandemic is bringing, I’m not optimistic that the next few years will be anything less than the further erosion of public education, but with this will come an opportunity to reshape it. I’ll be right alongside you, trying to realize that more just vision of public higher education.
If you are a UW student who was not accepted into Informatics recently, and you’re feeling desperate, don’t work through those feelings alone. Call UW SafeCampus if you’re in crisis, talk friends and family, and if you like, write me to share your thoughts and I will reply (eventually).