Months after DOGE gutted the federal Institute of Education Sciences (IES), Linda McMahon’s Department of Education is pivoting to the work of reforming it. That won’t be easy. DOGE reduced IES to rubble, laying off almost every IES staffer and canceling pretty much every contract while making little obvious effort to distinguish the good from the bad or the useful from the useless.

The upside is a chance to rebuild a more agile, cost-effective, valuable agency. But again, not easy. For one thing, a lot is still uncertain: many staff may be returning (due to a court injunction), some canceled contracts aren’t actually gone, and the department is hip-deep in high-stakes legal fights on multiple fronts. For another, given the chaos and the rush of deadlines, there’ll be a temptation to stitch things back together, pretend there’s been meaningful change, and call it a day. That would be a lost opportunity.

Where to start?

I find it useful to think of IES’s traditional activities as falling into three broad buckets: statistical collections, evaluations of tailored (and, thus, replicable) interventions, and, well, everything else. The “everything else” constituted the lion’s share of IES activity.

Given that, here are five thoughts on how best to reimagine IES’s role.

First, IES should prioritize collecting timely, reliable data on the warp and woof of American education, which is also the mission most closely aligned with the Constitution’s Article I Weights and Measures Clause. That means bolstering the National Assessment of Educational Progress, School and Staffing Survey, Integrated Postsecondary Education Data System, and the like. Right now, post-DOGE, these data collections are being held together by duct tape and skeleton crews. That’s a travesty. This is the work that only an appropriately funded and staffed IES can credibly do, and it provides the foundation for informed public debate and policymaking.

Second, IES should seize the excuse to leave all its other legacy initiatives on the cutting-room floor. Will that put lots of education evaluation shops and professors out of luck? Probably. So be it. Academe intent on dabbling in the familiar ed-school fascinations can still hit up philanthropy, find clients, or seek institutional support. The majority of federal education research funding is likely to be gobbled up by the tiny share of researchers who do this work responsibly. That’s fine. We often hear talk about needing a Flexner report to transform education research and training. Well, Flexner was largely about culling the quacks and crackpots to reshape the medical ecosystem. Here’s the chance to do that for education.

Third, refocusing IES can’t just be an excuse to put lipstick on the pig. There’ve been suggestions, for instance, that IES may be able to set things right if it simply adopts new research “priorities.” While there’s nothing overtly wrong with this, it risks becoming an escape hatch to avoid larger changes to the agency’s mission and direction. There’s nothing easier than for researchers to pay lip service to new priorities by tweaking any proposal to suit them. The upshot is that clearer, more academic priorities may help, but only as part of a larger shift.

Fourth, IES needs to bring in a new mix of researchers. A big problem with a focus on “priorities” rather than on the broader mission is that it’s likely to leave the same cast of academic authorities and external reviewers in positions of influence. Dynamic change requires revamping the review process so as to bring in more scientists and scholars from outside the familiar education research bubble.

Fifth, any restructuring should focus on tasks rather than bodies—a tip I discussed years ago in Cage-Busting Leadership. As former IES director Mark Schneider, has pointed out, IES was overrun by a culture of slack, turf war, and padded contracts. There’s a need to rethink structures, staff, workflow, and contract costs. The question shouldn’t be how many staff used to work on a given project but what tasks need to be done. How many hours a week are needed for contract oversight? Communications? Budgeting? Then figure out the skills and staff to meet those needs. Body counts steer us back to the familiar. Task analysis yields more productive conversations.

Thus far, DOGE’s “move fast and break things” ethos has certainly yielded a lot of breakage. Secretary McMahon’s team now has the opportunity to redeem the expectation that DOGE would be an agent not just of destruction but of reform. Here’s hoping they can deliver on it.

Frederick Hess is an executive editor of Education Next and the author of the blog “Old School with Rick Hess.”