## Reading layer `school_absences' from data source 
##   `https://drive.google.com/uc?export=download&id=10iHcg3hJcTfNHonzlUaEXiC5ob042WgU' 
##   using driver `GeoJSON'
## Simple feature collection with 2116 features and 5 fields
## Geometry type: POINT
## Dimension:     XY
## Bounding box:  xmin: 438439.7 ymin: 65253.22 xmax: 3029310 ymax: 1021744
## Projected CRS: NAD83 / North Carolina (ftUS)

Executive Summary

Chronic absenteeism means missing at least 10% of the school year. This matters because students who miss a lot of school lose instructional time and tend to have worse academic outcomes. Research has also linked frequent absence with weaker executive-function and social-emotional outcomes.

Main finding: In this sample, the average post-COVID chronic absenteeism rate is 27.8%, compared with 14.9% before COVID. That is an average increase of 12.9 percentage points.

Why Chronic Absenteeism Matters

Chronic absenteeism is more than occasional truancy. Students may miss school because of illness, transportation problems, mental health challenges, family responsibilities, housing instability, or disengagement. No matter the reason, the result is lost instructional time.

Keppens (2023) used administrative attendance data from more than 62,000 secondary students and found that unexcused absence, sickness absence, and school exclusion were all negatively associated with academic achievement. Ansari and Pianta (2022) also found that more frequent school absence was associated with lower academic, executive-function, and socio-emotional outcomes.

These findings show why chronic absenteeism is an important policy issue. Missing school can affect more than grades; it can also affect long-term student development and engagement.

Current Post-COVID Absenteeism Patterns

Post-COVID chronic absenteeism: descriptive statistics (%)
Schools Mean Median SD Minimum Maximum
2116 27.8 26.8 11.8 5 80.5

The average post-COVID chronic absenteeism rate is 27.8%, while the median is 26.8%. Looking at both values is useful because a small number of schools with especially high absenteeism could pull the mean upward.

The distribution shows that absenteeism is not evenly spread across schools. This is important because a statewide average can hide schools facing much more severe attendance problems.

Rural Versus Non-Rural Schools

Hypotheses

Null hypothesis (H₀): The mean post-COVID chronic absenteeism rate is the same for rural and non-rural schools.

Alternative hypothesis (H₁): The mean post-COVID chronic absenteeism rate differs between rural and non-rural schools.

Assumption Checks and Test Choice

The two groups are independent because each school belongs to only one rurality category. The outcome is continuous, and the group sizes are large. I also look at group summaries and boxplots for skew and extreme outliers.

Because the sample sizes are large and I do not want to assume equal variances, I use a Welch two-sample t-test.

Post-COVID chronic absenteeism by rurality (%)
rural_group n mean sd median
Non-rural 1193 26.5 11.9 25.5
Rural 923 29.4 11.4 28.5

The average post-COVID chronic absenteeism rate is 29.4% in rural schools and 26.5% in non-rural schools. The difference is 2.9 percentage points.

The Welch t-test gives t = -5.68 and p = <0.001.

Because p < .05, I reject the null hypothesis. There is statistically significant evidence that post-COVID chronic absenteeism differs between rural and non-rural schools.

From a policy perspective, even a difference of a few percentage points can represent many students. Rural schools may also face different barriers, including transportation problems, long travel distances, limited access to health care, and fewer local services. That means statewide solutions may need to be adjusted for local context.

High-Income Versus Low-Income Schools

Hypotheses

Null hypothesis (H₀): The mean post-COVID chronic absenteeism rate is the same for high-income and low-income schools.

Alternative hypothesis (H₁): The mean post-COVID chronic absenteeism rate differs between high-income and low-income schools.

Assumption Checks and Test Choice

The groups are independent because each school belongs to only one income category. The outcome is continuous, and both groups are large. I again inspect descriptive statistics and boxplots for unusual skew or extreme outliers.

I use a Welch two-sample t-test because it does not require equal variances.

Post-COVID chronic absenteeism by school income group (%)
income_group n mean sd median
High-income 1058 23.8 10.0 23.1
Low-income 1058 31.7 12.2 30.7

The average post-COVID chronic absenteeism rate is 31.7% in low-income schools and 23.8% in high-income schools. The difference is 7.9 percentage points.

The Welch t-test gives t = -16.35 and p = <0.001.

Because p < .05, I reject the null hypothesis. There is statistically significant evidence that post-COVID chronic absenteeism differs between high-income and low-income schools.

This comparison matters because poverty can create attendance barriers that schools cannot solve through punishment alone. Transportation instability, housing insecurity, health needs, caregiving responsibilities, and limited access to services can all make regular attendance harder.

Before Versus After COVID

Hypotheses

Null hypothesis (H₀): The mean within-school change in chronic absenteeism from pre-COVID to post-COVID is zero.

Alternative hypothesis (H₁): The mean within-school change in chronic absenteeism from pre-COVID to post-COVID is not zero.

Assumption Checks and Test Choice

This is a paired comparison because the same schools are measured before and after COVID. I therefore calculate the within-school change and use a paired t-test.

The important normality assumption applies to the distribution of the differences, not to the two time periods separately. Because the sample is large, the paired t-test is generally robust to moderate non-normality.

Before-versus-after COVID comparison (%)
n pre_mean post_mean mean_change sd_change median_change
2116 14.9 27.8 12.9 7.2 12.4

Across matched schools, the average chronic absenteeism rate changed from 14.9% before COVID to 27.8% after COVID. That is an average change of 12.9 percentage points.

The paired t-test gives t = 81.9 and p = <0.001.

Because p < .05, I reject the null hypothesis. The evidence indicates that average chronic absenteeism changed significantly after COVID.

The paired design is especially useful because each school serves as its own comparison. That reduces noise from stable differences between schools.

Policy Recommendation

Recommendation: North Carolina should fund a tiered, barrier-focused attendance strategy that uses early-warning data, personalized family outreach, and targeted practical supports instead of relying mainly on punishment.

The results suggest that chronic absenteeism should not be treated as one uniform problem. Differences across school types, along with the overall change after COVID, point to barriers that may vary across schools and communities.

A practical statewide strategy should include three parts:

1. Identify problems early.
Schools should use current attendance data to flag students before they cross the 10% chronic-absence threshold. Early intervention is more useful than waiting until a student has already missed weeks of school.

2. Find the real reason for absence.
Schools should work with families to determine whether absences are driven by transportation, physical or mental health, housing instability, caregiving responsibilities, bullying, disengagement, or another issue.

3. Target resources where the need is greatest.
Schools with persistently high absenteeism or sharp post-COVID increases should receive extra technical assistance and flexible resources. Support might include transportation help, school-based health services, family outreach, or partnerships with community organizations.

This recommendation is consistent with research showing that successful attendance strategies often address the barriers behind absence rather than treating attendance as only a compliance problem. Community-school approaches that combine family engagement, integrated supports, and a positive school climate may be especially useful.

Limitations

This analysis compares school-level averages and does not prove that rurality, school income, or COVID directly caused the differences that appear in the data. School-level absenteeism can also be influenced by grade configuration, local health conditions, transportation, student demographics, school climate, and community resources.

The hypothesis tests show whether differences are statistically detectable in this sample, but policy decisions should also consider the size of those differences, local context, and whether proposed solutions are realistic to implement.

Conclusion

Chronic absenteeism is a serious educational problem because it reduces students’ access to instruction and is associated with worse academic and developmental outcomes. This analysis examines post-COVID patterns across North Carolina public schools, compares rural and non-rural schools, compares high-income and low-income schools, and tests whether absenteeism changed significantly from before to after COVID.

The strongest policy response is not one statewide punishment or incentive. A better approach is to use timely data to identify problems early, engage families, address the actual barriers behind absences, and direct more support toward schools facing the greatest challenges.

References

Ansari, A., & Pianta, R. C. (2022). The grade-level and cumulative outcomes of absenteeism. Child Development, 93(4), e406–e420.

Keppens, G. (2023). School absenteeism and academic achievement: Does the reason for absence matter? Learning and Instruction, 86, 101769.

Learning Policy Institute. (2024). Reducing chronic absenteeism: Lessons from community schools.

North Carolina Department of Public Instruction. (2026). AttendNC Bright Spots: North Carolina’s attendance bright spots show what’s possible.