The Trump Administration has published its announcement about proposed changes to the admissions data the Department of Education collects through IPEDS. This announcement answers some of the questions I raised in a previous post about President Trump’s executive order. There’s actually a lot to like here in terms of increased transparency, but the scale and timing are, to use a technical term, nuts. I discuss the proposed changes below, but let’s cover some basics first so we can understand just how radical the scale and timing of these planned changes are.
What is IPEDS?
The Integrated Postsecondary Education Data Survey (IPEDS) is a survey that every institution of higher education that receives federal financial aid dollars is required to complete every year. It collects a huge amount of data across a broad range of topics. This data is used by institutions, policymakers, researchers, and advocates to understand higher education better and hold it accountable. IPEDS currently has 16 components (e.g., Institutional Characteristics, Financial Aid, Fall Enrollment) that almost 5,000 institutions fill out in three surveys that are completed in the fall, winter, and spring. Completing IPEDS is a very large task, typically handled by an office of institutional research, which has to coordinate with academic departments, administration, financial aid, admissions, and many other programs across a college or university to complete the survey every year.
What is in the Admissions Survey?
The Admissions survey component opens in December and closes in February. It asks colleges and universities a lot of questions, which you can see here but here are some that tend to be particularly useful:
- Admissions considerations (e.g., test scores, legacy status, essays)
- Applications, disaggregated by sex and race
- Admits, disaggregated by sex and race
- Enrollments, disaggregated by sex and race
- Test score submission percentages and score ranges
- Waitlist, early action, and early decision
Several of those components, such as the disaggregation by race on applicants and admits as well as data on early action and early decision numbers, are new this December. Some of those changes were proposed and vetted in a technical review panel in 2021 and underwent public comment periods. They were announced to the admissions field more than a year ago, and institutions saw the new survey questions in December 2024. That’s how changes are normally made to IPEDS. They take a long time to develop; guidance is issued; and institutions have time to prepare for the changes.
It’s very important to note that not all institutions need to complete all the questions on the admissions survey. If an institution uses an open admissions policy–as many four-year and almost all two-year and short term programs do–than it can skip many of the questions, so these proposed changes will likely affect between 1,500 and 2,000 institutions.
What Does the Trump Administration Want to Add to IPEDS
The Trump Administration wants to create a whole new Admissions and Consumer Transparency Supplement” (ACTS) survey component for IPEDS in 2025-26, i.e., the academic year that is kicking off now across the country. This component will include a lot of redundant questions that replicate items already in the Admissions component, and it will add many new questions that will increase transparency in college admissions.
It will be a huge amount of work and coordination for admissions, financial aid, and student outcomes offices and even more for the offices of institutional research that are responsible for reporting to IPEDS. Completing the proposed ACTS survey will be a big, but not impossible lift for flagship universities and wealthy private institutions that have larger offices for institutional research. There are, however, many smaller institutions required to fill out the Admissions survey that have institutional research offices run by a single person or that rely on part-time workers and that cannot afford the latest data management systems that make this kind of reporting more efficient. The increased administrative burden on them will be large, particularly at a moment when many insitutions are facing financial and staffing challenges.
Did I mention that the Trump Administration also wants the last five years of data for all the questions in this new survey component?
It’s not just the scale of the proposed changes that will increase the burden on universities; it’s the timing. It is not clear which of the three seasonal surveys the administration want to add this new admissions transparency component to, but they are all too soon. The fall survey opens next month and closes in October; the winter survey runs from December to February; spring runs from December to April. Unless the Department of Education creates a new data collection period, the best case-scenario is that colleges and universities will need to prepare and report on six years of data for new survey questions by mid-April.
Once again, in a normal IPEDS process, changes are made after a technical review panel of experts considers the changes and after comment periods (required by law) because it is important to create guidance for institutions to complete the survey, so they know, for instance, how to calculate GPAs or determine family income for people who never submitted a FAFSA. It is impossible to see how the National Center for Education Statistics (NCES), which administers IPEDS, can get this guidance to institutions, particularly since almost all NCES employees have been laid off.
Given the scale and timing, it will be nearly impossible to trust this first batch of data, which, again, will cover six years. Even if institutions make their best effort, there are bound to be mistakes at a level we have never seen before. That is reason enough to delay this process until 2026-27 at the soonest. Expect legal challenges, the very least, on the timing of the changes.
What is in the ACTS Survey Component?
Before getting into the weeds here, I want to call out the largest concern with this expanded data admissions collection: it creates significant opportunities for operators to weaponize admissions data against institutions in ways that are likely to create significant pressure to reduce racial and socioeconomic diversity on campus. (I get into more detail on this and other issues here.) Disaggregating so many components by race will inevitably lead to accusations of racial discrimination against institutions, particularly when many of the survey components are ill-defined and other important admissions priorities, such as legacy preferences, athletic recruitment, and relationships with feeder schools, are not going to be included in the data collection. Add to this the fact that there are serious problems with using aggregated data of this kind to determine the role that single factors like race played in admissions decisions and enrollments, even when you’re doing it in good faith.
So yes, we can expect the data to collected through this new component to be used irresponsibly and in bad faith, but I do not think that undercuts the value of transparency. It makes it more important. Bad arguments should be answered with better ones, and better ones require better data. This announcement is better than what I feared it would be, which was defining transparency as disaggregating by race, test scores, and GPA. Institutions, researchers, and advocates have been seeking much of this data for a long time, and it is on us to make sure that we meet every bad argument with a better one. And, yes, the polarized state of our politics and media make it even harder for sophisticated arguments to be heard, let alone listened to, but that’s still no argument for letting admissions operate in the shadows.
Here are the proposed changes (all language in italics is from the announcement):
For undergraduate students, we anticipate the [ACTS] component will collect data by race-sex pair on
- the count of institutions’ applied, admitted, and enrolled cohorts, both overall and further disaggregated by admission test score quintiles, GPA quintiles, ranges of family income, Pell Grant eligibility, and parental education.
- the average high school grade point average and admission test score quintiles for institutions’ applied, admitted, and enrolled cohorts
- the count of students admitted via early action, early decision, or regular admissions.
IPEDS has been collecting data on applicants, admits, and enrollments by sex for years. This year it will begin collecting applicant and admit data by race and ethnicity for the first time; it already did so for enrollments. Institutions are prepared for this change, although this announcement does not say anything about the small number of institutions that do not include questions about race/ethnicity on their application.
Disaggregating applicants, admits, and enrollments by all these other categories is entirely new and will come with significant questions and challenges. Breaking these questions down by race and sex will create some subsets that are so small that the data will have to be suppressed for privacy reasons, especially with early decision pools. For example, Harvard enrolled 3 freshmen in 2024 who identified as Native American or Alaska Native. If you disaggregate a pool that small by race and/or sex, it would take little effort to identify the exact students with those scores, leaving aside the meaningless of creating quintiles for 3 people. These privacy issues with small subsets are an issue for all the new questions.
- Test score quintiles
- Colleges currently report the 25th, 50th, and 75th percentile scores of their enrolled class, so asking them to make that the 20th, 40th, 60th, and 80th percentile is not a big lift, but every little change in process is a change that a human being must make.
- Most colleges in the United States are test optional, so those who report tend to have higher scores, which means that data reporting on test scores has become less reliable, as scores keep creeping up.
- GPA quintiles
- GPAs are not at all standardized across high schools. It is certainly not the case that all high schools report grades on a 4.0 scale. Many high schools weight grades, so it’s not unusual to see 4. 4 and 5.7 GPAs. Some colleges unweight GPAs in their calculation; others do not. Some include all coursework; others only include academic subjects. Before GPA quintiles can become useful data points, NCES needs to create guidance on how to calculate them.
- Incomes ranges and Pell Grant eligibility
- Having a better sense of the income status of who applies to, gets admitted to, and enrolls in a college, disaggregated by race and sex, would be very interesting and useful.
- The problem is that colleges do not have that information for most applicants and admits. That income data and Pell eligibility come through the FAFSA, and not all students will submit a FAFSA to a college before they are admitted. Some will never submit one. Even with enrollments, the data will be partial. Many enrolled students at some institutions will never submit a FAFSA because they can pay the full cost of attendance.
- Parental education
- As with income data, it will be helpful to see the highest level of education that an applicant’s parents attained and how that factor plays into who applies as much as who is admitted and enrolls.
- The problem here is whether a college has that data. The Common App does include a question about parent’s highest level of education attainment in their general survey, which provides some level of standardization, but there are hundreds of institutions that do not use the Common App. They may not collect that data, or they may collect it using different categories.
- Among newly enrolled undergraduate students, we anticipate the ACTS component will collect data by race-sex pair on both the count and average amount of students receiving:
- any institutional grant aid
- merit-based institutional grant aid
- need-based institutional grant aid and
- any local, state, or federal government aid overall,
This will be further disaggregated by admission test score quintiles, GPA quintiles, ranges of family income, and enrollment via early action, early decision, or regular admissions.
These new questions have the same issues as the previous ones, but I think that they are important to include in the data and could be very revealing about just who gets the most institutional grant aid and merit aid in particular. Reporting this data at such a granular level will be new and represent an increased burden on institutions.
- We anticipate the ACTS component will also collect data overall and by race-sex pair on (1) students’ average cumulative GPA at the end of the academic year; (2) the average cost of attendance, and further disaggregated by admission test score quintiles, ranges of high school grade point average, ranges of family income, and enrollment via early action, early decision, or regular admissions. (3) graduation rates further disaggregated by admission test score quintiles and ranges of high school grade point average; and (4) graduates’ final cumulative grade point average. Additional data may be gathered to better understand remedial or other non-credit coursework for newly enrolled students.
This may be the biggest ask of them all. The Department of Education is looking for outcomes data at a very granular level. I suspect not many institutions are collecting this data currently, and it will take a lot of work to do so. As with the financial aid data, I think it is good data to collect and publish, but it is wholly unreasonable to expect institutions to report it this coming academic year.
- The survey will cover undergraduate and graduate admissions.
The inclusion of graduate admissions is the big news here. The announcement is vague on what questions will be asked about graduate admissions, but it does indicate that “graduate programs are to be broken down by fields of study.” This is a very large undertaking. For undergraduate admissions, the institutional research office had to coordinate with one office–admissions. Getting all this data from every graduate program is going to be a massive task.
There is a lot more to come on these proposed changes. Expect a lawsuit or lawsuits. Even more than I did last week, I have a hard time seeing how these changes will be implemented this year, but I also do not think the changes are going to go away. What’s important now is getting the changes made responsibly, in consultation with stakeholders and experts, in a reasonable time frame so colleges and universities can actually fill out this new IPEDS component in a way that will shed light rather than shadows on college admissions.
And, oh yeah, it sure would be great to see legacy admissions added to the data collection.
Leave a Reply