The push for no-zero grading policies, which prohibit assigning zeros for class assignments or tests, began during a broader “equitable grading” movement arising from the view that traditional grading is rooted in bias and discourages students. Critics counter that such policies inflate grades in ways that demotivate students and that higher standards produce better long-term outcomes.
This debate extends beyond policy circles. Multiple Fordham Institute studies show that teachers and students share concerns about no-zero grading policies. Eighty-one percent of teachers say giving partial credit for assignments that are never turned in harms academic engagement. Similarly, high school students report that weak grading standards, along with widespread AI-driven cheating, diminish their learning experience. Yet despite such findings—and a dearth of empirical evidence supporting the practice—no-zero grading continues to spread. About 27 percent of teachers nationally report working in a school with some form of no-zero policy.
In a working paper, Dan Goldhaber, Maia Goodman Young, and Luke Caldwell at the National Center for Analysis of Longitudinal Data in Education Research (CALDER) aim to map out the no-zero grading policy landscape in one state. They examined 295 traditional school districts in Washington State, covering 2015–16 through 2024–25, and collected documents ranging from board policies to school-level student handbooks.
Using AI[1], the team created a system[2] to determine what counts as a no-zero policy and identified 68 districts with such policies. To better understand what types of districts are more likely to adopt no-zero policies, the CALDER team estimated the probability of adoption in a given year, controlling for district size, location, test proficiency rates, as well as student demographics. They also looked at variation across schools within the same district.
The study finds that no-zero grading policies increased by 110 percent over ten years across the school districts examined. These policies are most common at the elementary level, but high schools had the fastest growth rate: 350 percent since 2016. Many of these no-zero policies predate the Covid-19 pandemic, with nearly half of the elementary-level policies’ adoption dates predating the 2015–16 school year.
The results also reveal enormous variation in how districts write and implement no-zero policies. Sixteen percent of those with such policies have contradictions between board policy and school handbooks. For example, one source established a no-zero policy while another permitted zeros. There is a particularly wide range when it comes to enforcement. Some districts require students to redo their assignment with mandatory structured support sessions, while others exclude the assignment from the gradebook. Some districts assign a partial score, while some policies vary by assignment type.
Perhaps the most striking finding is an equity paradox. While these reforms are typically framed as equity tools, Washington districts with more low-income students were less likely[3] to have a no-zero policy. At the elementary level, districts serving more special education students were also less likely to have no-zero policies. Among high schools, suburban districts were more likely to adopt such policies than city districts of similar size and student population.
The study is limited, of course—to a single state and to published policy documents. It cannot capture how teachers implement these rules in the classroom. Yet it exposes the unevenness of no-zero grading policies, which are hard to find and often inconsistent even within the same district. This may favor privileged students in multiple ways. Families with more education and institutional savvy may, for example, be better positioned to advocate for their children when rules are unevenly applied. And because high-poverty districts have fewer no-zero policies to begin with, this gap could give wealthier students a GPA advantage in college admissions. No-zero policies are framed as equity tools, yet their distribution doesn't match their intent. The landscape seems most uneven for the students these policies were designed to benefit.
SOURCE: Dan Goldhaber, Maia Goodman Young, and Luke Caldwell, "Where Have All the Zeros Gone? The Landscape of No Zeros Grading Policies," CALDER Working Paper No. 345-0826 (August 2026).
[1] The method they used is called retrieval-augmented generation. It uses large language models to search and classify documents, grounding the response in verified source material. The LLM’s response is pulled only from the source material rather than from the LLM’s training data. The CALDER team completed this process using two LLMs: Sonnet 4.6 and Claude Opus 4.7.
[2] It was a “typology system” that examined the following in each policy document: the source of the policy, what the policy applies to, how zeros are structurally eliminated, and how missing or plagiarized work is handled.
[3] A 10 percentage point increase in a district’s low-income student share is associated with a 0.93 percentage point decrease in the likelihood of no-zero policy adoption.
