When the COVID‑19 pandemic led to the closure of school buildings across Georgia in March 2020, the Georgia Policy Labs (GPL) and our metro-Atlanta district partners were already several years into a research-practice partnership. Because the legal agreements, data pipelines, and personal relationships were in place, we quickly pivoted from our existing research agenda to answer urgent questions about lost learning time, remote instruction, and recovery strategies. Five years and more than a dozen studies later, five key takeaways stand out for researchers, policymakers, practitioners, and funders planning for the next disruption.

Build the house before the storm

Districts needed actionable evidence while decisions were still on the table throughout the pandemic, not two years later. We initially signed data‑sharing agreements with our district partners in 2018, and by 2020 the process to securely transfer data from partners to GPL was routine. We quickly realized that having access to formative assessment data was key to tracking achievement almost in real time. This access is how we documented steep early setbacks in math, especially for students experiencing poverty, and the uneven rebound that followed. 

What should researchers, policymakers, practitioners, and funders do now?

• Formalize partnerships early. Establish memoranda of understanding and secure data-sharing agreements before a crisis hits.
• Protect a common baseline. Encourage partner districts to collect nationally normed assessments (e.g., i‑Ready or MAP Growth) at least twice—and ideally three times—a year. These assessments formed the backbone of every GPL estimate of pandemic-related achievement growth.
• Document implementation. Tracking which schools lengthened intervention blocks or offered small‑group tutoring allowed us to evaluate what actually happened inside buildings, not just what districts had planned.
• Support rapid response research. Pivoting to answer urgent questions amid a disruption requires flexible funding to cover data preparation, research, and dissemination.

Autonomy with guardrails beats total decentralization

When recovery dollars arrived, one district let each principal design a schedule and choose interventions, resulting in wide variation, little alignment with evidence, poor record keeping, and mixed achievement gains. Schools that devoted more minutes to structured intervention blocks did post modest growth in ELA and math, but ad‑hoc tutoring—often in large groups after school—showed no clear benefit. 

What’s the implication for the next disruption? 

• Flexibility, but not a free-for-all. Give school leaders room to tailor supports, but only after central offices narrow the menu to strategies with a solid research base and set minimum quality standards (e.g., group size and frequency for tutoring). Total autonomy can scatter resources and delay recovery.
• Be selective about who gets what. Spreading resources too thinly can dilute their impact. Districts should define clear eligibility criteria for interventions—and stick to them. In several cases, we uncovered that students who didn’t meet the original eligibility criteria were included based on discretion or informal recommendations. While well-intentioned, these changes made it harder to measure impact and may have weakened program effectiveness.

Small natural experiments are research gold

Snow‑day Zoom lessons, pollution‑related closures, and short‑term virtual academies are now part of K‑12 life. Each episode is a natural experiment that researchers can analyze to help students. Districts that archive log‑in records, assignment data, and attendance during these blips create a living lab for studying remote‑learning tools without another global pandemic. 

What opportunities lie ahead? 

• Remote‑day playbooks. By comparing achievement on “digital‑learning days” to nearby in‑person days, researchers can test which platforms (synchronous, asynchronous, or blended) and which lesson designs keep students on track.
• Targeted tutor pilots. Short, opt‑out tutoring programs embedded in the school day could reveal the dosage and staffing models that are effective before districts commit to district‑wide roll‑outs.
• Blend methods to explain what numbers alone reveal. Follow up natural experiment findings with interviews, focus groups, or classroom observations—especially when data expose gaps. Qualitative insights can uncover why a strategy works for some students but not others and point to refinements before the next disruption.

Strengthen evaluation systems before you need them

Our studies leaned heavily on districts that already used i‑Ready or MAP Growth across grades. Systems without consistent formative data could not see where students stood until state tests came back—months too late to adjust.

• Adopt a common formative assessment (at least fall and spring) that yields national percentiles, allowing rapid comparisons across years and subgroups.
• Collaborate with partners to streamline data flows so researchers receive cleaned files within days, enabling a rapid response.
• Create good record-keeping systems, so that when strategies to get students back on track are rolled out, teachers and program administrators know how and where to input implementation data in a clear and consistent fashion. Knowing who participated and what the selection criteria were is essential for causal analysis of program impacts.

Translate evidence into plain guidance 

From our full portfolio, three cross‑cutting messages resonate with both classrooms and boardrooms: 

• Focus dollars on in‑school, high‑dosage supports. Voluntary, after‑hours programs rarely reach the students who need the most help.
• Match teachers to the medium. Veteran educators often excel online; some star in‑person teachers struggle behind a webcam. Use data on past virtual performance when staffing remote classes.
• Tie relief funds to accountability. Linking dollars to transparent goals nudges schools toward evidence‑based interventions.

Looking forward 

Educational disruptions—whether a hurricane, a cyber‑attack, or another public‑health emergency—are inevitable. They do not have to stop learning. By investing now in resilient RPP structures, baseline assessments, and thoughtful governance, researchers, policymakers, and practitioners can turn the next crisis into a source of insight to bolster student supports, rather than a setback. Disruptions are inevitable, but it’s how we prepare that matters. 

Please visit our website to learn more about our COVID-19 response. Follow us on LinkedIn and subscribe to our newsletter to keep up with GPL’s latest research.

Photo by Allison Shelley/The Verbatim Agency for EDUimages