Addressing high school dropout rates starting at the elementary school level

The City Connects program is an initiative of Boston College that works to address non-cognitive barriers to student success among elementary school pupils in Boston Public Schools (BPS), as well as charter and private school students in Boston and other nearby cities. It was piloted in six low-performing BPS elementary schools in 2001, assessing needs and providing access to services for students via a third party rather than through the schools themselves.

Those services can include academic tutoring, social-emotional development, health needs, and family supports. Full-time coordinators are embedded in the schools and monitor need, referrals, and successful use of these services, obviating the need for teachers to become de facto social workers and for school administrators to become service providers. A fuller description of the program can be found here. It is important to note, for purposes of this study, that City Connects is limited to elementary school.

Because of its genesis within Boston College’s Lynch School of Education, City Connects has been widely studied by the school’s researchers. The latest report looks at long-term effects of the program on combatting high school dropout. The students under study were part of the first five cohorts of kindergarteners and first-graders to participate in the pilot program, from the school years 2000–01 through 2004–05. A team of researchers led by Terrance Lee-St. John followed the enrollment histories of those youngsters throughout their schooling in BPS through the 2013–14 school year, when the youngest of the students were in ninth grade and the oldest had just graduated high school. The comparison group comprised students in the same school-year cohorts who attended BPS schools not participating in City Connects. Students were excluded from both groups if they received high school instruction in “substantially separate” special education placement, if they permanently transferred out of BPS prior to reaching ninth grade, or did not reach ninth grade by 2013–14 due to being held back in prior years. The final sample studied included 894 City Connects students and 10,200 non-City Connects students.

Because of the small number of school buildings involved in the pilot initially, all treatment effects were calculated at the student level. Overall, treatment students registered a 9.2 percent dropout rate in high school, compared to 16.6 percent for the non-treatment students, a fairly wide variance that points to significant positive effects, especially for an intervention that happened years prior. The most significant benefits showed up for black students, those who qualified for free or reduced-price lunch, and males. Interestingly, BPS provides withdrawal codes indicating the reasons—as far as they are known—for students dropping out. While more treatment students than non-treatment students dropped out due to incarceration and for unexplained reasons, more treatment students also dropped out to take up employment or due to GED completion than did their non-treatment peers.

Caveats offered by Lee-St. John and his team are worth noting. Selection effects could threaten the validity of the study if, for example, more conscientious and involved families enrolled in the treatment schools, which could drive a spurious correlation between the program and student outcomes. Since researchers cannot control for the families’ conscientiousness, we should take these large effect sizes with the caveat that selection bias could be driving some of the findings. Additionally, the observable demographic makeup of students in the treatment and non-treatment schools is somewhat different, although the researchers describe efforts to control for those differences.

How does an elementary-level intervention help reduce the likelihood of dropouts by nearly fifty percent years after treatment has stopped? Part of the reason could be that students are connected with service providers independent of their elementary schools. It is possible that service provision along many lines continued throughout middle and high school, but the researchers do not test that possibility, noting only that the effects of treatment are significant over time. City Connects touts its multi-faceted and student-centric support structure as addressing multiple barriers to student success. This and other research from Boston College seem to point to its effectiveness, but the aforementioned caveats give rise to some skepticism, and more data would help confirm these effects and pinpoint the source.

SOURCE: Terrance J. Lee-St. John et. al., “The Long-Term Impact of Systemic Student Support in Elementary School: Reducing High School Dropout,” AERA Open (October 2018).

 
 
Jeff Murray
Jeff Murray is the Ohio Operations Manager of the Thomas B. Fordham Institute,