New Haven Public Schools (NHPS) serves approximately 20,000 K-12 students across a choice system of traditional, magnet, and charter schools. But choice systems are not inherently fair or transparent: approximately 1 in 2 families submit applications that only give them a 30% chance of admission or less. Over the past several years, TetherEd and NHPS have partnered to build enrollment infrastructure grounded in the same peer-reviewed research that first diagnosed where school choice systems fall short for lower-income families. The result is a choice process that is more transparent, better supported, and more likely to connect every family with a school that meets their needs. Multi-year results show meaningful growth in family engagement and placement outcomes, with field-tested feedback interventions reducing non-placement risk by nearly half.

New Haven Public Schools has offered families a choice among traditional, magnet, and charter schools for decades—making it a well-studied urban choice district, and the direct subject of the peer-reviewed research that underpins TetherEd's tools. Its student population is predominantly lower-income, and the promise of school choice, that families could find the right fit regardless of where they lived, was real in principle. In practice, the enrollment process itself was often getting in the way.

Professor Christopher Neilson and his colleagues were the first to document precisely how. Their research showed that the mechanism New Haven used to assign students rewarded families who knew how to game it and systematically disadvantaged those who didn't—and that the families least equipped to navigate that complexity were mostly lower-income. When those findings reached district leadership, New Haven acted on them. The district overhauled its assignment mechanism based directly on the research. After implementing the new assignment mechanism, the strategic gap across families shrunk dramatically.

Fixing the assignment mechanism resolved one layer of the problem. But as the research deepened, it became clear that the mechanism was only part of the puzzle. Even in a strategy-proof system, families still had to find schools, understand the process, and submit informed rank-order lists—and the evidence showed they were struggling with all three. Lower-income households in New Haven faced the largest information burdens and were least likely to seek out supplemental guidance on their own. Families routinely overestimated their odds at popular schools, submitted short lists, and ended up unmatched through no fault of the mechanism.

The same patterns emerged when Neilson and collaborators studied school choice markets in Chile at national scale: information gaps, misestimated placement odds, and concentrated risk among the families with the fewest fallback options. It was this convergence of findings, across two very different contexts, that pointed toward a broader set of tools. Fixing the rules of the game mattered. So did helping families play it.

That body of work became the foundation for TetherEd. The problems it set out to solve were not hypothetical. They were documented, quantified, and reproducible across systems. The stakeholder burden broke down like this:

• For lower-income families in New Haven: Families who ranked schools by genuine preference rather than strategic calculation regularly went unmatched. In the most recent cycle, nearly half of all applicants submitted initial rank-order lists carrying a 30% or greater non-assignment risk, and a meaningful share had virtually no chance of placement.
• For schools: Opaque demand signals made enrollment projections unreliable, generating phantom seat demand that complicated planning and staffing.

For the district: Problems only became visible after placement had already occurred—too late for any intervention to matter.

TetherEd's tools were built to address each layer of the problem the research had identified — not as a generic product suite, but as direct translations of documented failures into working infrastructure.

• Algorithm Design. The foundation was already in place: TetherEd supported NHPS's operation of a fully strategy-proof Deferred Acceptance matching algorithm, ensuring families could rank schools by genuine preference without strategic penalty. But as the research had shown, the algorithm alone wasn't enough.
• Virtual School Explorer. Families couldn't make informed choices about schools they didn't know existed. The Explorer gives every NHPS family a single place to browse school profiles, photos, videos, and historical admissions data, reducing the search costs that the research identified as a primary driver of short, risky rank-order lists. Virtual tours and immersive school profiles extend the district's visibility to families who would not otherwise encounter those schools during the application window.
• Application Feedback Suite. The most direct translation of research into a product. A 2022 paper in the Quarterly Journal of Economics by Arteaga, Kapor, Neilson, and Zimmerman had documented that personalized, real-time placement warnings — delivered mid-application to families at non-assignment risk — reduced non-placement risk in New Haven by nearly half. The Feedback Suite operationalizes exactly that intervention at scale, flagging at-risk applicants and recommending safer additions before the window closes.
• AI Virtual Assistant and Outreach Suite. Multilingual automated guidance that extends enrollment support beyond the portal itself — reaching families less likely to seek help proactively, acting as a trusted extension of the district’s support office during peak season.

Each addition targeted a failure mode the evidence had already named. District engagement deepened at each stage—what began as leadership acting on pre-publication findings became a sustained partnership, with NHPS teams using TetherEd's data layer to monitor application behavior in real time and intervene earlier in the cycle than had previously been possible.

Across multiple cycles, tool engagement has grown year-over-year and is producing measurable changes in application behavior. 

• For families: Engagement with the Application Feedback Suite is growing year-over-year, and it is translating into action. Thousands of families receive personalized feedback reports with live admission probabilities and recommendations for alternative schools based on their preferences and lottery characteristics. Hundreds of families adjusted their applications as a result of the feedback. In previous years, the equivalent of four classrooms full of students would have gone without a match in New Haven.
• For schools: Under the new algorithm, schools receive demand signals that reflect genuine family preferences rather than strategic gaming. Leveraging the content-rich School Explorer, they are able to showcase what makes them unique and fight against misconceptions or low visibility. Schools with a strong Explorer presence saw significantly more family engagement and applications than those without.
• For the district: The combination of a strategy-proof mechanism and active guidance tools has given NHPS a real-time view of application behavior that did not exist under the prior system, shifting the district from reactive problem-solving after placement to early intervention during the choice window. Each family that enrolls in a district school rather than exiting to a private or charter option outside the system represents meaningful revenue—in Connecticut, this pushes $20,000 per student annually.

‍

What can other districts learn from New Haven? Here are a few key lessons:

1. The equity costs of enrollment process design are often invisible—until you look for them.
Most districts with school choice programs collect outcome data after placement, not during. That means the families most affected by information gaps (those who submitted high-risk rank-order lists and went unmatched) often aren't visible in post-cycle reporting at all. They simply don't appear in the next year's applicant pool. The New Haven research found these dynamics to be measurable, consistent, and concentrated among lower-income households. Districts that track enrollment outcomes by income level, not just aggregate placement rates, tend to find a gap. Whether the tools to address it are in place before the next cycle is a separate question, but identifying the gap is the necessary first step.

2. When information gaps go unaddressed, families don't just go unmatched. They leave.
The New Haven evidence is consistent with what Neilson and collaborators found across national-scale systems: lower-income households carry the largest information burdens and are the least likely to navigate a complex process without support. Outreach that requires families to seek it out (e.g., a school fair, an FAQ page, an optional information session) tends to reach families already engaged with the process. Families at risk of going unmatched rarely seek help; they need to be met mid-application, before the window closes. When those families exit to private or out-of-system options instead, the consequences compound. The information gap is an equity problem, but it is also a budget one.

3. Tools built from field research come with a baseline that internally developed solutions don't.
The feedback intervention underlying TetherEd's Application Feedback Suite was rigorously evaluated in New Haven before it was a product. That publication record doesn't eliminate the need for local implementation work, but it does change what a district is taking on when it deploys the tools. The question shifts from "will this work?" to "how do we make it work here?" That is a meaningfully different starting point.

TetherEd is dedicated to partnering with education agencies, schools, and communities to advance transparent, equitable, and effective enrollment processes. Whether you are a government agency, school district, or education provider looking to modernize your enrollment infrastructure, we would love to connect. Reach out to zack@tether.education.

‍