A welcome trend in recent decades has been the sustained rise in high school graduation rates up to a national average of 85% for 2017-18—as well as a significant narrowing of differences by race and ethnicity for this educational milestone (though gaps remain).
Policymakers are rightly focused on making sure even more young Americans successfully complete their high school education, and on further narrowing gaps between various subgroups. To that end, the Department of Education requires states to report high school completion rates for the prior academic year to track progress at a national level.
Former Intern - The Brookings Institution
States are also required to provide these data for a variety of subgroups, including “each major racial and ethnic group,” economically disadvantaged students, children with disabilities, and English learners. Following the passage of Every Student Succeeds Act (ESSA) in 2015, states are also required to report disaggregated data on high school graduation rates for homeless students, and for those in foster care; though this data is not yet being reliably reported.
This disaggregated data has proven valuable for assessing progress towards more equitable outcomes, especially for marginalized groups. But there is one glaring omission in the subgroups for which data is available: sex. We do not know the national high school graduation rates for girls and boys, since states are not required to provide this data—but, we argue here, this requirement should be added.
Across the US, fewer boys are graduating
Reliable high school graduation rates by sex are not available at a national level, but, because states are required to track individual students’ graduation status across other demographic groups, many states already collect and publish the rate by gender. (Note that state Departments of Education often report rates by sex, rather than gender identity.) To gauge the national trend, we collected publicly available high school graduation data for the 2016-17, 2017-18, and 2018-19 school years.
In the 2017-18 school year—the school year we collected the most data for—we accounted for over three quarters of the graduating cohort across 37 states. We use the preferred measure, the Adjusted Cohort Graduation Rate (ACGR); see the technical note below for more details on measurement. The pattern was clear: while the graduation levels and gender gap vary by state, in every state for which we collected data, girls graduated at a higher rate than boys:
Stark gender gap in national high school graduation rates
Pooling the data from these 37 states, we construct an estimate of the national graduate rates by sex. In 2018, about 88% of girls graduated on time compared to 82% of boys—a 6 percentage point gap. By contrast, the gap between the graduation rate among white students (89%) and the graduation rate among Black students (79%) is 10 points, and the gap between Hispanic students and white students is 8 points (89% v. 81%). The graduation rate for boys is only slightly higher than for economically disadvantaged students (82% v. 80%):
Our data suggests, then, that the gender gap is sizable. Assuming that the national gender gap in high school completion is the same as in our sample of 37 states, over 45,000 fewer US boys than girls would have graduated high school in the 2017-18 school year. By comparison, around 23,000 young adults age out of foster care each year.
Our analysis also shows a significant variation in high school graduation rates across states. In the 2017-2018 school year, the female high school graduation rate was 9 points higher than the male rate in Mississippi, the state with the largest gender gap in high school completion that year. In North Dakota, the state with the smallest gender gap, the graduation rate among girls was only 2 points higher than among boys.
Hispanic and Black boys fare the worst
These gender gaps intersect with race gaps, with especially poor outcomes for Black and Hispanic boys. Unfortunately, disaggregated data by race and sex is not available on the most up-to-date and preferred measure, the ACGR. But it is available for a less robust metric, the AFGR (see technical note)—which shows much bigger gender gaps for Black and Hispanic students. In the 2012-13 school year, for example, there was a 4 point gender gap in high school graduation among white and among Asian or Pacific Islander students, compared to an 8 point gap among Hispanic students and 10 points among Black students. States should disaggregate their data by both race and sex for the ACGR in order to track these trends more accurately.
These gaps in high school graduation rates prefigure similar differences in college enrollment. For example, 2018 college enrollment among Hispanic women was 9 percentage points higher than that of Hispanic men, and Black women’s enrollment was 8 percentage points higher than that of Black men.
Get the data
The state-level data we collected shows an unmistakable gap in high school graduation rates for boys and girls. There is also evidence that race and socioeconomic status impact students differently by gender—in particular, boys are more likely to be negatively impacted by low socioeconomic status.
The federal government should follow the lead of the dozens of states already collecting and publishing graduation rates by gender. The gender gap is certainly wide enough, and important enough, to justify adding sex to the list of subgroups for which states are required to report data to the federal government, and to ask states to provide disaggregated data by sex by race. This is the whole reason for asking for disaggregated data: to track trends for subgroups. Gender gaps in education matter, in either direction.
Specifically, the Department of Education should issue guidance for states to report their Adjusted Cohort Graduation Rates by sex, and to report these rates by sex for economically disadvantaged students, and for each major racial and ethnic group. Perhaps this is an issue that will be taken up by the new Commission on the Social Status of Black Men and Boys?
Technical Note: How is high school graduation measured?
- Averaged Freshman Graduation Rate (AFGR). The AFGR divides the number of diplomas granted in a year by the average enrollment across the corresponding 8th, 9th, and 10th grade classes. For example, the 2016-17 AFGR is the total number of diplomas granted at the end of the 2016-17 school year divided by (the number of 8th graders who enrolled in fall 2012 + the number of 9th graders who enrolled in fall 2013 + the number of 10th graders who enrolled in fall 2014) divided by three.
- Adjusted Cohort Graduation Rate (ACGR). The ACGR, by contrast, divides the number of cohort members who earned a diploma within four years by the number of first-time 9th graders four years earlier, adjusted for student transfers, emigration, and death.
Because the AFGR does not account for cohort changes, it is sensitive to migration, and may mismatch students’ starting cohort if they graduated early or late. Similarly, because students may drop out during their freshman year and the AFGR averages 8th, 9th, and 10th grade starting enrollment, the AFGR likely underestimates total enrollment; 9th grade dropouts should be counted as first-time 9th graders, but they are excluded from 10th grade enrollment. This is especially troubling when it comes to estimating boys’ graduation rates—they are more likely than girls to drop out of school, so the AFGR likely overestimates their graduation rate (and the overall high school completion rate). Although it includes estimates by gender, a 2011 report by the National Research Council suggests that aggregate measurements like the AFGR are “not sufficiently accurate for research, policy, or accountability decisions” and should no longer be used. In the 2010-11 school year, the Department of Education began requiring states to report the ACGR. For policymakers concerned with gaps among these groups, the ACGR provides more reliable (although still imperfect) estimates by tracking individual students.
"The pandemic has highlighted just how intricately related lack of broadband access is to systemic inequality."