DC’s Undercount of Students Designated At-Risk
The number of students designated as “at-risk” are declining in the District, but should it be?
The first school year the District implemented a new “at-risk” definition to support a more equitable student funding model was in 2014-15. At-risk students, those whose family receive SNAP or TANF benefits, are homeless or in the foster care system, or are overage for their grade in high school, receive additional weight in the school funding formula. These categories were chosen based on the differences in academic outcomes (DC CAS proficiency back then, high school graduation, etc.) between these students and their peer group.
Then, in 2016, the DC Council passed the Fair Shot Minimum Wage Amendment Act to progressively increase the minimum wage to $15 an hour by 2020, representing a 58% increase in the minimum wage since 2014. Meanwhile, the thresholds for SNAP and TANF, both tied to federal poverty levels that generally annually increase at the rate of inflation, have only increased 10% over the same timeframe, leaving many families earning an income that exceeds thresholds that qualify them for governmental benefits yet still below DC “living wage” thresholds that incorporate local costs for housing, food, childcare, medical, transportation, and other typical expenses.
While the District has experienced ongoing gentrification, we find one key contributing factor to the decline in the total number of students designated as at-risk is the minimum wage increase. This (rather lengthy) post examines the relationship between District population shifts, minimum wage hikes, the number of students who are designated as at-risk, and the likely undercount of students living in households that can benefit from the additional academic and social-emotional resources that the “at-risk weight” was designed to equitably fund. Given the disparate impact of the pandemic on the outcomes for students designated as at-risk, we find the likely undercount of such students to be relevant to recovery funding conversations.
Changes in overall and student population counts since the implementation of at-risk designation
One of the best data resources for examining comprehensive changes in District population and shifts in family earnings is through the annual American Community Survey (ACS) conducted by the U.S. Census Bureau. Individual deidentified household data from the ACS is made available for public research purposes (the IPUMS dataset) with the most recent published dataset collected in 2019. Since the at-risk designation includes multiple components, we utilize data showing student counts by at-risk component that OSSE provided before the last two Council oversight hearings as well as overall at-risk numbers from audited enrollment files.
DC public school enrollment and overall population figures from OSSE and ACS
Overall, the total number of at-risk students declined 1,400 students over the last five years even though the total PK-12 public school enrollment has increased by 8,000 students.
The chart above shows trends in overall population and student enrollment as well as those qualifying for the various at-risk factors between 2014 and 2019. It also shows general population and students attending schools East of the River (EotR) in some of the city’s least gentrified and lowest-income neighborhoods. From 2014 to 2019, we find a significant increase in the number of homeless students, limited change in the number of students in the foster care system or who are overage in high school, and significant drops in students whose family qualifies for SNAP or TANF governmental benefits.
In Wards 7 and 8, using data from the Census’s annual American Community Survey (ACS), we find the total school-aged population living in a household (either a single-parent or multi-parent household with children) not earning a living wage increased by approximately 3,600 students between 2014 and 2019. Yet, the number of at-risk students in schools located East of the River declined by 500 students.
Why are the at-risk numbers declining East of the River if the number of families below the living wage has increased?
Back when the at-risk weight first went into effect in 2014, the DC minimum wage was $9.50, and a family of four with two parents working an average of 32 hours a week at minimum wage jobs earned $2,432 monthly, under the DC SNAP threshold of $2,584 in 2014. A similar family of four with two minimum wage job earners working an average of 32 hours per week took home $3,840 a month in 2019, well above the SNAP threshold of $2,839 that year.
However, our example family’s disposable income to support their children’s social-emotional and academic progress did not increase as dramatically now that they are fully responsible for food costs. The Urban Institute published a study showing the limited overall impact on disposable income families working minimum wage jobs would have available after having to compensate for qualifying for fewer governmental benefits. The family is still far from earning a “living wage” in the District of Columbia, which was a monthly income of $4,822 in 2019 based on MIT’s calculations.
This example “nuclear family” of four is atypical of families East of the River where 55% of households are led by a single parent. Only 24% of single-parent EotR families earned enough to meet the living wage threshold based on the number of kids in the household.
In addition to not earning a living wage, minimum and near-minimum wage workers are significantly more likely to work hours that are atypical of the 9-5 Monday through Friday of salaried workers earning higher wages, leaving them with less time to meet the social-emotional and academic needs of their students outside of school. The federal Bureau of Labor Statistics’ data shows earners in the bottom quartile are more than twice as likely to work hours outside the regular daytime and two and half times more likely to work weekends than the top half of all earners. We believe these families, while no longer meeting the technical definition of “at-risk,” benefit from additional student supports at school.
How large is the at-risk undercount and what are the financial ramifications for schools?
DC is very likely undercounting the number of students who are most at risk of academic and social-emotional hardship as explained above. We take three different approaches to estimating the undercount’s range and financial impact.
Approach 1: At-risk components as a general proxy for a larger population
The original adequate funding study from 2013 discussed a new “at-risk” definition as a replacement for the previous free or reduced-price lunch economically disadvantaged student definition for funding purposes. A new measurement was needed because of the implementation of the federal Community Eligibility provision that allowed all students to get free meals when the school served high enough levels of impoverished students, reducing paperwork and financial transaction burdens. The new “at-risk” definition, mostly derived from the number of families qualifying for SNAP or TANF benefits, essentially leaves out those “reduced-price meal” families who do not earn a living wage and cannot qualify for SNAP benefits. The 2013 proposed at-risk weight of 0.37 was designed to be a large enough proxy to help cover the difference in no longer counting these families.
As we have noted above, the total number of students in families living below the living wage threshold has not significantly changed, an estimated 53,039 in 2014 to 53,045 in 2019, since the introduction of at-risk. Meanwhile, the total number of students qualifying as at-risk via SNAP has dropped from 36,572 to 35,359. The at-risk weight as a proxy for all low-income students approach would suggest an undercount of about 1,200 students or approximately $3.4M for schools based on the proposed FY22 UPSFF base rate and minor at-risk weight increase.
Approach 2: Approximating the number of students below the living wage not receiving SNAP/TANF
Since the proposed at-risk weight of 0.24 is significantly lower than the 2013 recommended weight of 0.37, the budgeted amount is not large enough to fully serve as that intended “proxy” role and cover all students in families not earning a living wage. Approach #2 uses data from the Census’s annual American Community Survey sample; though, the best data source will be the detailed 2020 Census information which will be released later this year (only total population counts for Congressional reapportionment considerations and racial demographics have been released so far).
We find that between 27% and 37% of DC households whose income exceeds the SNAP threshold did not earn a living wage in 2019. If we consider all the students in this category to meet the intent of being covered by the at-risk designation’s funding weight, then the current at-risk undercount would be 16,500 students with a 90% confidence range between 14,000 and 19,000 students. With the current difference between proposed at-risk weight and a fully funded weight of 0.13, this undercount would translate to an underfunding of about $18M next year alone.
Approach 3: Using a sample of actual student data to calculate at-risk exiters
We analyzed changes in at-risk student designations for 3,176 students who were continuously enrolled at 13 of our partner elementary and middle schools over the last three school years to estimate how many students have exited at-risk status since the implementation of Fair Shot minimum wage increases. From that data, we found that 188 out of the 1,466 continuously enrolled students (12.8%) who were designated at-risk in October 2018 had lost that designation by October 2020.
Assuming this trend held true across the city during the years since the launch of the at-risk designation that were impacted by Fair Shot, we estimate the total number for currently enrolled students who exited at-risk status due to no longer qualifying for SNAP/TANF benefits to be about 9,400 students. This leads to an estimated total at-risk underfunding using FY22 proposed weights of $26.2M.
What are our recommendations for next steps?
To be clear, we believe SNAP and TANF counts represent our best, most reliable set of validated household income information. Likely, there is a strong correlation in the percent of students designated at-risk at schools and the percent of the student population whose families do not earn a living wage. The at-risk designation was meant to be an equity-centric proxy to better ensure the schools that were more likely to need additional resources to best serve their students receive the funding to acquire those resources.
There are several immediate options to mitigate the impact of the at-risk student undercount and equitably ensure schools serving more families who are likely to need additional social-emotional and academic support have the funding:
Utilize American Rescue Plan (ARP) general funds (approximately $2.3 billion) coming to the District to support social-emotional and academic interventions that will predominantly benefit students who are designated as at-risk or part of the undercount as noted above;
Increase the at-risk weight from 0.24 (current proposed budget from Mayor) to at least 0.30, about half the difference between the current weight and the recommended weight from the 2013 Adequacy study; and/or
Redefine the universe of “at-risk students” to be those students who have met the definition anytime within the last three years. This can help keep more students whose parents no longer qualify for SNAP but are not earning a living wage stay in the funding category.
Before FY23 budget discussions begin less than a year from now, it may make sense to dig into the Census data to more accurately capture how many students’ families are ineligible for SNAP/TANF benefit but do not earn a living wage, adjusting the recommended at-risk funding weight accordingly.