The Trump administration released further details on its order for colleges to supply more racially disaggregated admissions data and wants to hear from the public about its plan.

A draft of the proposal, which will officially be published Friday on the Federal Register, states that certain institutions will be required to collect and report comprehensive data about their admissions decisions going back five years. It must be broken down by race and sex and include students’ high school GPA, test scores, time of application (early decision, early access or regular decision) and financial aid status, among other things.

However, the new survey component, which the Department of Education is calling the Admissions and Consumer Transparency Supplement, will not affect all colleges and universities—just four-year institutions that use “selective college admissions,” as they “have an elevated risk of noncompliance with the civil rights laws,” officials wrote in the notice.

(The document does not say anything about reporting data on legacy admissions, another practice that, like affirmative action, has received public pushback in recent years.)

Members of the public will have 60 days to comment on the notice. Among other things, the department wants feedback on what institutions should be subject to the new reporting requirements as well as the anticipated burden the request will place on university staff.

Some higher education scholars and officials are already chiming in with their concerns informally.

University of Tennessee higher education professor Robert Kelchen wrote in a post on LinkedIn that not only will the request be a “substantial lift” for colleges, but also for staff at the department who run the Integrated Postsecondary Education Data System and will manage the data on the back end.

The Department of Education laid off nearly half its staff—including most of the employees at the National Center for Education Statistics, which would collect and analyze the data—in March.

“I’d love to see the survey form where all of this data would be collected—because after years of sitting in [meetings] where we figured these things out, the sheer number of variables/elements and the lack of any definition around the vagueness of them demonstrates the loss of the knowledgeable NCES staff they lost,” wrote Carolyn Mata, a consultant who works in institutional research, in a response to Kelchen’s post. “This is a case of throwing everything possible at the wall.”