The Reckoning / Part 2 of 3

The Phantom Present

When teachers don't record attendance, eSchool marks every student present. This investigation maps the gap between what we measure and what we report.

Finding
KBC reports the highest attendance rate in the KIPP Delta district at 96.9% YTD. It also has the lowest attendance compliance at 26% — meaning nearly three-quarters of all expected attendance checks were never performed. On days when almost no one records attendance, the system defaults every student to “present.” The number we have been celebrating is, in significant part, a measurement artifact.
KBC
26%
YTD Compliance
Reported Attendance: 96.9%
DCHS
59%
YTD Compliance
Reported Attendance: 93.8%
DELA
75%
YTD Compliance
Reported Attendance: 94.6%
BCPS
91%
YTD Compliance
Reported Attendance: 95.7%
01 The Inverse Relationship
Schools with lower compliance report higher attendance. The charts below overlay both metrics week by week. Watch how KBC's compliance hugs the floor while its attendance floats near the ceiling.
Weekly Attendance Compliance Rate (% of periods where attendance was recorded)
Weekly Reported Student Attendance Rate

What you are seeing

BCPS maintains ~91% compliance and reports ~96% attendance. Their number is trustworthy — teachers are actually recording who shows up. KBC averages 26% compliance and reports 97% attendance. Most of that 97% is the system auto-filling “present” because nobody logged the real data.

The pattern is consistent: the less you measure, the better the numbers look. This is not anyone gaming the system. It is a structural incentive problem — the default value is the happy answer.

How Phantom Attendance Works

eSchool, like most student information systems, requires a daily attendance record for every enrolled student. When a teacher submits attendance for their class, students marked absent get an “A” code. But when no one submits attendance at all, the system needs a value — and the default is “present.”

Teacher submits attendance Real data recorded Trustworthy
No one submits attendance System defaults to present Phantom

This means low compliance does not just create missing data — it creates false positive data. Every unchecked student inflates the attendance rate. A school could have 50% real attendance, but if nobody records it, the dashboard shows 100%.

02 Day-Level Forensics
Each dot is one school day. X-axis is compliance, Y-axis is reported attendance. At BCPS, dots cluster high on both axes — real measurement, real attendance. At KBC, dots jam to the left with attendance suspiciously high regardless of how little was measured.

The tell

On their lowest-compliance days, KBC and DCHS report attendance approaching 100%. That is not stellar teaching — that is the system defaulting every unchecked student to present.

BCPS barely has any dots below 75% compliance. They are doing the work. Their attendance data can be trusted as a genuine measure of who was in the building.

03 Compliance Heatmap
Every bar is one school day, colored by compliance rate. Green means measured. Red means a gap in the data.
04 Exhibit A: Phantom Days at KBC
The most extreme examples — days where almost no attendance was recorded, yet the system reports near-perfect attendance. These are not anomalies. They are the norm.

February 4, 2026

Compliance: 2.3%. Of the roughly 90 expected attendance checks at KBC that day, two were completed. Reported attendance: 100.0%. Every single student marked present — not because someone verified they were in their seats, but because the system filled in the blanks.

Compare to BCPS on the same day: 96.2% compliance (they actually checked), 100.0% attendance. Both report the same outcome, but one is a measurement and the other is a default.

05 What This Means for the District
KBC is the starkest case, but this is not a single-school problem. DCHS runs at 59% compliance. Here is what the numbers look like when we estimate real attendance from only the records that were actually measured.

The arithmetic

When compliance is low, the reported attendance rate is a blend of real data and defaults. We can estimate the “real” rate by looking only at students whose attendance was actually recorded. If unchecked students follow the same pattern, the district-wide attendance picture changes substantially.

At KBC, the estimated real rate is roughly 88% — nearly 9 points below the reported 96.9%. At DCHS, the gap is about 4 points. BCPS, with 91% compliance, barely moves. The lesson: compliance is not an administrative checkbox. It determines whether we are seeing the truth.

Next: The Cost of Not Knowing

When we cannot measure who is actually in the building, we cannot identify who is about to leave. Part 3 of The Reckoning quantifies what that blindness costs — in dollars, in students, and in trust.