promise worth keeping

8/9/2019 Promise Worth Keeping

1/24

A PROMISEWORTH KEEPING: Advancing the High School Graduation Rate in Philadelphia

Julia Ransom, Ph.D.Heather Grifs, Ph.D.Jennifer Eder, M.P.H.

Vaughan Byrnes, M.Sc.Benjamin French, Ph.D.Robert Balfanz, Ph.D.

Douglas Mac Iver, Ph.D.David Rubin, M.D., M.S.C.E


2/24

2 Keeping Our Promise

ABOUT THE AUT HORS:

Julia Ransom, Ph.D.Research Associate, PolicyLab,

The Children's Hospital of Philadelphia Heather Grifs, Ph.D.Research Scientist, PolicyLab,The Children's Hospital of Philadelphia Jennifer Eder, M.P.H.Policy and Research Associate, PolicyLab,The Children's Hospital of Philadelphia Vaughan Byrnes, M.Sc.Research Associate, Everyone Graduates Center,

School of Education, Johns Hopkins University Benjamin French, Ph.D.Department of Biostatistics and Epidemiology,University of Pennsylvania Robert Balfanz, Ph.D.Co-Director, Everyone Graduates Center,School of Education, Johns Hopkins University Douglas Mac Iver, Ph.D.Principal Research Scientist, Everyone Graduates Center,School of Education, Johns Hopkins University David Rubin, M.D., M.S.C.E.Professor of Pediatrics, Perelman School of Medicineat the University of PennsylvaniaCo-Director, PolicyLab,The Children's Hospital of Philadelphia


3/24

A Promise Worth Keeping 3

EXECUTIVE SUMMARYGraduating from high school is a critical milestone for youth achieving a successful

transition to adulthood. Youth without high school diplomas often face few joboptions, low annual incomes, and poor health outcomes. High dropout rates within acommunity are related to higher poverty and crime rates, less tax revenue, and moremoney and resources spent on social services.

In 2006, Unfullled Promise: The Dimensions andCharacteristics of Philadelphia’s Dropout Crisis, 2000–2005 shed light on how many students drop out ofhigh school in the School District of Philadelphia(SDP) and what factors put students at risk for

dropping out. The results were grim: for studentsentering high school between 1997 and 2001, theon-time graduation rate hovered around 50 percent.Risk factors associated with dropping out includedchronic absence, child welfare involvement, andfailing grades. After the release of Unfullled Promise ,several citywide initiatives (including accelerated highschool programs, afternoon and evening classes,re-engagement efforts, and workforce developmentand occupational skills training) were expanded orcreated to address the dropout crisis.

In 2014, Project U-Turn commissioned PolicyLab atThe Children’s Hospital of Philadelphia and JohnsHopkins University to conduct a follow-up study toexamine whether the prevention and interventioninitiatives reduced dropout rates by addressing the

following questions:

• What were the cohort graduation and dropoutrates? What factors related to these rates?

• How did subpopulations vary in graduation ratesand risk factors for dropout?

• What were the re-engagement rates for highschool dropouts over time? Of the students whodropped out of high school and re-engaged,how many participated in a Project U-Turn re-engagement program?

• How many students eventually enrolled in post-secondary institutions?


4/24

4 A Promise Worth Keeping

Graduation rates increased and dropout rates havedecreased since the rst Unfullled Promise report.

• Four percent fewer students dropped out of highschool during the 2011–2012 school year than duringthe 2003–2004 school year.

• The 2008–2009 cohort graduation rate reached a highof 64 percent, an increase of 12 percentage pointsfrom the previous report’s highest graduation rate.

• The four-year cohort dropout rate decreased from 29percent for the 2003-2004 cohort to 25 percent for the2008–2009 cohort.

• Grade promotions and attendance rates increased asgraduation rates increased.

Graduation rates remain lower for at-risk students,

including those involved with the city’s Department ofHuman Services (DHS), black and Hispanic males, andfemales who gave birth prior to and/or during highschool.

• Graduation rates for students with DHS involvementcontinued to lag behind those of their non-involvedpeers, with juvenile justice-involved youth having thelowest graduation rate.

• Hispanic males and females made the greatest gainsin graduation rates, increasing 18 percentage pointsand 15 percentage points, respectively, between the2002–2003 and 2008–2009 cohorts.

• Graduation rates increased for black and Hispanicmale students, but they remain lower than thoseof their white and Asian male counterparts and allfemales.

• Graduation rates for adolescent mothers remainedlow, peaking at 43 percent with the 2007–2008 cohort.

• Students involved with juvenile justice and adolescentmothers had the highest probability of dropping out ofhigh school of all high-risk groups examined.

Of the students who dropped out of high school, anincreasingly large percentage re-engaged in either theschool system or a re-engagement program.

• 54 percent of dropouts in the 2008–2009 cohort re-engaged, up from 47 percent in the 2002–2003 cohort.

• The graduation rate for dropouts who re-engagedremained steady at roughly 35 percent across theseven cohorts observed. This suggests that while re-engagement programs pulled more dropouts back in,they may not have provided easier or more effectiveways of achieving a high school diploma.

• One in ve students across all cohorts, and over half ofall re-engaging dropouts, participated in at least oneProject U-Turn program: GED to College, EducationalOptions Program, E 3 Centers, Accelerated HighSchool, Gateway to College, and Occupational Skills

Training Program.

Although graduation rates increased, rates ofenrollment in two-year and four-year post-secondaryinstitutions failed to keep pace.

• Enrollment in post-secondary institutions for allstudents marginally increased.

• Students involved in DHS and adolescent motherscontinue to lag far behind in post-secondaryenrollment, and students in juvenile justice enroll in

post-secondary education at substantially lower ratesthan all other DHS categories.

CONCLUSION

Overall, more students are graduating from Philadelphiapublic schools since the release of the rst UnfullledPromise report. However, several groups continueto lag behind in terms of graduation and dropoutrates. Providing additional dropout prevention andinterventions for the most at-risk students (those withDHS involvement, adolescent mothers, and black andHispanic males) may help them complete high school.Additionally, more students who drop out are re-engaging, and a majority of those are participating in oneof the programs implemented or expanded as a result ofProject U-Turn.

K E Y F I N D I N G S


5/24


6/24


Questions addressed in this report include:

• What were the cohort graduation and dropout rates?What factors relate to these rates?

• How did subpopulations vary in graduation rates and

risk factors for dropout?

• What were the re-engagement rates for high schooldropouts over time? Of the students who dropped outof high school and re-engaged, how many participatedin a Project U-Turn re-engagement program?

• How many students eventually enrolled in post-secondary institutions?

To provide context for the questions addressed in thisstudy and the landscape of dropout intervention effortsfollowing the release of Unfullled Promise , the nextsection provides background information regardingthe formation of Project U-Turn and its initiatives fordisconnected youth.

PROJECT U-TURN

Unfullled Promise documented the multidimensionalproblem of Philadelphia’s dropout crisis and emphasizedthat its resolution would require action on many fronts.Project U-Turn partners have worked collectively overthe last eight years to raise public awareness about the

dropout issue, expand educational options for strugglingstudents and youth who have left school without adiploma, support research and measure results, 5 andleverage funds to support these ef forts.

While its efforts are multifaceted, Project U-Turn hasfocused on reconnecting young people to a wider arrayof pathways leading to a secondary credential andpostsecondary education or training. In fact, the launchof Project U-Turn featured a “come back fair,” whereformer dropouts could meet with leaders from areaschools, organizations, and programs to get back on

track educationally.

Many of the re-engagement and alternative educationoptions are coordinated through SDP’s Ofce of MultiplePathways to Graduation. For example, SDP contractswith providers to operate its Accelerated High Schoolprogram, which provides educational programs andservices for s truggling students and out-of-schoolyouth. Furthermore, in May 2008, SDP opened a Re-Engagement Center for youth seeking to return toeducation. Launched with seed funding from ProjectU-Turn, the Re-Engagement Center is based at the districtheadquarters, receives space, personnel, and operatingfunds from SDP, and has received stafng supportfrom the City of Philadelphia. The Center is a one-stopresource where former students can get informationabout various educational program options and bereferred to programs that meet their needs.


7/24


Examples of other pathways available for out-of-schoolyouth include GED to College and Occupational SkillsTraining programs, funded by federal WorkforceInvestment Act dollars and approved by PhiladelphiaWorks, Inc. E 3 Centers are supported by the TemporaryAssistance for Needy Families Youth DevelopmentFund and are made available through the Pennsylvania

Department of Human Services and approved byPhiladelphia Works, and by major annual investmentsfrom DHS. All of these pathways are competitivelyprocured and are operated by providers contracted byPhiladelphia Works and PYN.

In addition to expanding alternative education programsfor Philadelphia’s youth, Project U-Turn and its partnershave aggressively pursued policy changes and increasedfunding to advance dropout prevention and recovery.Key examples of these efforts include:

• Working with and supporting the School ReformCommission’s efforts to expand the number ofaccelerated schools and increase high-quality optionsfor off-track students and out-of-school youth.

• Leading advocacy efforts in 2011 and 2014 with locallegislators and stakeholders to maintain supportfor Accelerated High Schools when scal crisesthreatened their existence.

• Collaborating with the juvenile justice system andSDP to change education policies and practicesat delinquent placement facilities (for example,curriculum alignment, credit transfer rules, andcareer and technical education offerings) so thatthe academic and occupational coursework youthcomplete while in placement is more likely to becounted when students return to school distric tprogramming.

• Launching the DHS’s Education Support Center in2009 to improve educational stability and outcomesfor youth in DHS care, including children who are infoster care, receiving in-home services, or involved inthe juvenile justice system.

• Negotiating a memorandum of understanding tofacilitate the exchange of data between DHS and SDPregarding agency-involved youth and their progressin school.

• Commissioning and supporting a series of studies toassess the needs and realities of Philadelphia’s mostvulnerable populations (many of whom participate inthe aforementioned alternative education offerings)to inform program-based interventions and educationpolicy decisions.

• Working with school district ofcials to align studentmaternity leave policies with the state’s guidelines and

timelines so that teens can access child care subsidieswhen they return to school.

Project U-Turn and the programs initiated followingUnfullled Promise were created to stem the tide ofthe dropout crisis in Philadelphia. The next sectionsof the report provide analysis of graduation, dropout,re-engagement, and post-secondary enrollment in theyears following the release of that report.


8/24


Graduation rates increased and dropout rates decreasedsince the rst Unfullled Promise report.

ANNUA L DROPOUT FOR THE 20 11-20 12 SC HOOL YEAR

KEY FINDING: Four percent fewer students dropped out of highschool during the 2011–2012 school year than duringthe 2003–2004 school year.

Unfullled Promise provided a one-year snapshot ofdropout for the 2003–2004 school year for high schoolstudents, including those in charter schools. Thissection examines the magnitude of dropout for one

year in Philadelphia for all high school students enrolledat any point in the 2011–2012 school year, includingthose in charter schools. 6 See Table 1 for characteristicsof students in ninth through twelf th grade duringthe 2011–2012 school year. Of the 58,166 students, 6percent had limited English prociency, 16 percentreceived special education services, 19 percent wereinvolved with child welfare (including juvenile justice),and 5 percent of female students had given birth priorto or during high school.

Six percent of students dropped out in the 2011–2012

school year, and the percentage was relatively constantacross grades, ranging between 5 and 6 percent. Thisis a relative decrease of 4 percentages points from theannual dropout rates in the rst Unfullled Promise study,which were 10 percent for all high school students andranged between 8 and 13 percent for ninth throughtwelfth grades for the 2003–2004 school year (Figure 1).

Figure 1. DROPOUTBY GRADE forthe 2003–2004and 2011–2012School Years

Note: Data for both years include students enrolled in charter schools.Data for 2003–2004 obtained from Unfullled Promise (Neild & Balfanz,2006)

Graduation rates increased over the same time period, asreected by the cohort graduation rates detailed in thenext section.

0% 2% 4% 6% 8% 10% 12% 14%

12TH

11TH

10TH

9TH

2011-2012

2003-2004

PERCENT DROPOUT

G R A D E

8%

5%

6%

6%

6%

13%

11%

8%

What were the cohort graduation and dropout rates? What factors relate to these rates?

Table 1. CHARACTERISTICS OF ALL 9TH–12TH GRADE STUDENTS IN THE 2011–2012 SCHOOL YEAR

Note: Includes charter students. Characteristic s are for percentage of all students in column one, followed by percentage ofstudents with those characteristic s in each racial group.

Maternity

DHS Involvement

Special Education

Limited EnglishProciency 6% 19% 26%2%

2% 1%

1%

16% 15% 17% 17% 4%

19% 10% 24% 15% 6%

5% 6% 5%

All students:

58,166

White students:

7,989

Black students:

35,314

Hispanic students:

9,978

Asian students:

4,103


9/24


COHORT GRADUATION AND DROPOUT RATES

KEY FINDINGS:• The 2008–2009 cohort graduation rate reached a high

of 64 percent, an increase of 12 percentage pointsfrom the previous report’s highest cohort graduationrate.

• The four-year cohort dropout rate decreased from 29percent in the 2003–2004 cohort to 25 percent in the2008–2009 cohort.

• 3,170 students in the 2008–2009 cohort dropped out

of high school, compared to 5,493 in the 2002–2003cohort, which graduated before the implementationof Project U-Turn. This represents a decrease of over2,000 eventual dropouts.

• Grade promotions and attendance rates increased asgraduation rates increased.

This section examines seven cohorts of rst-time ninth-grade students from the 2002–2003 through 2008–2009 school years, followed up to four years from thebeginning of ninth grade. While an annual graduationor dropout rate gives an idea of how many studentsgraduate or drop in a single year, the cohort calculationprovides graduation and dropout information for groupsof students who start high school at the same time andare tracked over a given amount of time.

For example, a group of rst-time ninth graders canbe followed over four years to calculate the on-timegraduation rate for that cohort. The cohort rate can alsotrack a particular cohort for more than four years to allowstudents who dropped out more time to return andeventually graduate. 7

On-Time Graduation RatesThe percentage of students graduating in four years roseconsistently between the 2002–2003 and 2008–2009

cohorts, especially when compared to cohorts fromthe Unfullled Promise report. For example, Figure 2shows that the four-year graduation rate for studentsstood at 64 percent for the 2008–2009 cohort, while onlyone ninth-grade cohort had a graduation rate above50 percent between 1997 and 2001. 8 This increase isconsistent with the rise in graduation rates at the nationallevel. According to the National Center for EducationStatistics, the on-time graduation rate for rst-time ninthgraders in the 2008–2009 cohort reached a for ty-yearhigh of 80 percent. 9 While students in SDP lag behind thenational average, this is a substantial increase comparedto previous years.

Figure 2.FOUR-YEAR GRADUATION RATES OF 9TH-GRADE COHORTS

Note: Includes rst-time ninth graders beginning in SchoolDistrict of Philadelphia (SDP) schools (accelerated, specialadmission, citywide, and neighborhood). These ratesexclude ninth graders starting in charter schools, as well asstudents attending charters after the start of ninth grade.Data for 1996–1997 through 2000–2001 cohorts wereobtained from Unfullled Promise (Neild & Balfanz, 2006).The 2002–2003 through 2008–2009 cohort graduation datawere provided by SDP, and were calculated using SDP’sformer four-year cohort graduation methodology. Studentstransferring out of SDP without returning to SDP are notincluded. The dashed line marks a distinction between datafrom the original study cohorts and the cohorts included inthis follow-up study.

1997-2001Unfullled Promise

P E R C E N T G R A D U A T E D

2002-2009School District Reported

0

10

20

30

40

5060

70

2 0 0 8

- 0 9

2 0 0 7

- 0 8

2 0 0 6

- 0 7

2 0 0 5

- 0 6

2 0 0 4

- 0 5

2 0 0 3

- 0 4

2 0 0 2

- 0 3

2 0 0 1

- 0 2

2 0 0 0

- 0 1

1 9 9 9

- 0 0

1 9 9 8

- 9 9

1 9 9 7

- 9 8

48 % 44 % 48 % 43 % 52 % 52 % 53 % 57 % 56 % 58 % 61 % 64 %

COHORT YEAR


10/24


Dropout RatesThe four-year dropout rate ranged from 27 to 35 percentbetween the 1996–1997 and 2001–2002 cohorts. 10 Asgraduation rates increased following the release ofUnfullled Promise , dropout rates steadily declined.The four-year dropout rate reached a low of 25 percentamong the most recent cohort in this study (thosestudents who entered ninth grade in 2008–2009).Between the 2002–2003 cohort, which graduated prior tothe launch of Project U-Turn, and the 2008–2009 cohort,the number of high school dropouts decreased by over2,000 students, or by 42 percent (Figure 3). 11

The increase in graduation rates and decrease in dropoutrates were accompanied by other advances in studentachievement.

Graduation and AchievementThe increased graduation rates reected for recentcohorts were preceded by an improvement in severalindicators of progress toward graduation, such ashigher rates of on-time promotion to tenth grade andreductions in the percentage of students with chronicabsences. Philadelphia’s students have improved inseveral indicators related to graduation, includingchronic absenteeism, suspensions, and grade promotion.When a student experiences chronic absence, it canresult in loss of classroom time, which can impactlearning and lead to dropping out. 12 Chronic absence is

commonly dened as attendance of under 90 percent,the equivalent of missing a month or more of school. 13 Suspension has also been associated with an increasedprobability of dropout and a decreased probability ofgraduating. 14 The percentage of students experiencingchronic absence in eighth and ninth grades decreasedbetween the 2002–2003 and 2008–2009 cohorts (Figure4). Chronic absences for ninth-grade students rose inboth the 2005–2006 and 2008–2009 cohorts; however,they remain lower overall. Suspensions in ninth-gradedecreased from 37 percent in the 2002–2003 schoolyear to 24 percent in the 2008–2009 school year. The

rate of on-time promotion to tenth grade grew for eachsubsequent year and reached 81 percent by the 2008–2009 cohort.

Figure 4.ON- AND OFF-TRACK HIGH SCHOOL INDICATORS

Note: Indicators are not portrayed for course credits and failing grades

due to incomplete data within the s tudent-level dataset provided by theSchool District of Philadelphia.

Figure 3.FOUR-YEAR DROPOUT RATES OF 9TH-GRADE COHORTS

Note: Data were provided by the School District of Philadelphia (SDP),and were calculated using SDP’s former four-year cohort dropoutmethodology. This methodology included students whose rst schoolof ninth-grade attendance was an SDP or alternative school; it attributedstudents to their rst school of ninth-grade at tendance. Students whotransferred in or whose rst school of ninth-grade at tendance was acharter school are excluded from the calculations.

P E R C E N T D R O P O U T

COHORT YEAR

0

5%

10%

15%

20%

25%

30%

35%

2008-092007-082006-072005-062004-052003-04

29 % 27 % 31 % 31 % 31 % 25 %

010%

20%30%40%50%

60%70%

80%90%

100%

Suspended in 9th GradeChronic Absence, 8th Grade

Chronic Absence, 9th GradeOn-Time Promotion to 10th Grade

2008-092007-082006-072005-062004-052003-042002-03

COHORT YEAR


11/24


Graduation rates remain lower for at-risk students, including those involved with DHS, black and Hispanic males, and females who gave birth prior toand/or during high school.

RACE & GEND ER

KEY FINDING: Graduation rates increased for black and Hispanic malestudents, but they remain below those of their white andAsian male counterparts and all females.

To assess the progress made by specic racial andethnic groups over one year, Figure 5 shows the dropoutpercentages for white, black, Hispanic, and Asianstudents for the 2003–2004 and 2011–2012 school years.When compared to 2003–2004, the dropout rates for2011–2012 decreased for all races. However, gaps amongraces persist, with approximately 6 percent of blackand white students and 7 percent of Hispanic studentsdropping out in 2011–2012, compared to 3 percent ofAsian students.

Figure 5.DROPOUT RATES BY RACE/ETHNICITY IN THE 2003–2004

AND 2011–2012 SCHOOL YE ARS

Note: Data for both years include students enrolled in charter schools.Data for 2003–2004 obtained from Unfullled Promise (Neild & Balfanz,2006).

While the disparity in graduation rates among differentracial groups has improved since the publication ofUnfullled Promise , a graduation gap persists acrossracial/ethnic categories for students in Philadelphiaschools.

Figure 6.FOUR-YEAR GRADUATION RATES BY RACE/ETHNICITY

Cohort graduation rates increased for both males andfemales across all four racial/ethnic groups over time(Figures 7 and 8), although female students from mostracial/ethnic groups showed less improvement whencompared to their male counterparts. Despite lessimprovement compared to males, all female racial/ethnic

groups continued to outpace their male counterparts byas much as 8 percentage points for Asian females and11 percentage points for black females in the 2008–2009cohort.

0% 2% 4% 6% 8% 10% 1 2% 1 4%

HISPANIC

BLACK

WHITE

ASIAN

2 0 1 1

- 2 0 1 2

2 0 0 3

- 2 0 0 4

10%

11%

12%

7%

6%

6%

7%

3%

PERCENT DROPOUT

RACE/ETHNICITY:

How did subpopulations vary in graduation rates and risk factors for dropout?

0

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2008-092007-082006-072005-062004-052003-042002-03

ASIAN

HISPANICBLACK

WHITE P E R C E N T G R A D U A T E D

COHORT YEAR


12/24


The greatest gains in graduation rates were made amongHispanic males and females (increasing 18 percentagepoints and 15 percentage points, respectively, betweenthe 2002–2003 and 2008–2009 cohorts). The graduationrates for black and Hispanic males (currently at 58percent and 55 percent respectively) remain below thoseof their white and Asian male counterparts.

STUDENTS WITH DHS INVOLVEMENT

KEY FINDING: Graduation rates for students with DHS involvementcontinued to lag behind those of their non-involvedpeers, with juvenile justice– involved youth having thelowest graduation rate.

In a school environment where one in ve high schoolstudents is currently or was previously involved withthe child welfare and/or juvenile justice system, it isimportant to consider their potentially unique risks tohigh school completion compared to their non-involvedpeers. Figure 9 shows students with DHS involvement

had gradual improvements in graduation across alllevels of involvement in the child welfare system. Thegraduation rate for youth with foster care involvementincreased from 28 percent in the 2002–2003 cohort to44 percent in the 2008–2009 cohort, and the graduationrate for children receiving other DHS services increasedfrom 32 percent in the 2002–2003 cohort to 48 percent inthe 2008–2009 cohort.

The graduation rate for juvenile justice–involved youthexhibited the greatest gain among high-risk students,increasing from 16 percent in the 2002–2003 cohort to 36percent in the 2008–2009 cohort. Despite the gains forstudents with DHS involvement, their graduation ratesremain lower than their non DHS-involved counterparts. 15

Previous research in Philadelphia suggests that thiscould be due to higher rates of school change, delaysin enrollment, and a history of special education and/or behavioral problems. 16 However, it is worth notingthe gains seen for students with DHS involvement werelarger than those of their non-involved peers across thesame cohort years.

The progress made in graduation rates since 2002-2003by students with DHS involvement and those in juvenile justice may reec t increased information sharing andcollaboration between SDP and DHS through the 2008federal Fostering Connections to Success and IncreasingAdoptions Act and increased support through theDHS Education Support Center, which was establishedin 2010. Fostering Connections required that child

welfare agencies and school districts work together toensure that students do not experience disruptions inattendance and enrollment during changes in livingarrangements as the result of foster care.

Figure 7.FOUR-YEAR GRADUATION RATESFOR MALE STUDENTS

Figure 8.FOUR-YEAR GRADUATION RATESFOR FEMALE STUDENTS


COHORT YEAR

0

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2008-092007-082006-072005-062004-052003-042002-03

ASIAN

HISPANIC

BLACK

WHITE P E R C E N T G R A D U A T E D

COHORT YEAR

0

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2008-092007-082006-072005-062004-052003-042002-03

ASIAN

HISPANIC

BLACK

WHITE


13/24


ADOLESCENT MOTHERS

KEY FINDING: Graduation rates for adolescent mothers remained low,peaking at 43 percent with the 2007–2008 cohort.

Adolescent mothers who gave birth prior to and/or during high school had modest improvements in

graduation rates, increasing from 31 percent (2002–2003cohort) to 42 percent (2008–2009 cohort). Femalestudents who did not give birth during this time showedsmaller gains, from 66 percent (2002–2003 cohort) to72 percent (2008–2009 cohort). Among all high-riskstudents, adolescent mothers experienced the leastimprovement in graduation rates since the UnfullledPromise report.

PROBABILITY OF DROPOUTFOR HIGH-RISK STUDENTS

KEY FINDING: Students involved with juvenile justice and adolescentmothers had the highest probability of dropping out ofhigh school of all high-risk groups examined.

A large body of literature identies the risk factorsassociated with dropping out, including belonging to anethnic minority group, having limited English prociency,and receiving special education services. 17 Researchalso shows that adolescent mothers and students withchild welfare involvement, specically foster care and juvenile justice, are at a higher risk of dropping out. 18 Thissection explores how strongly these risk fac tors relateto students’ likelihood of dropping out of high school inPhiladelphia. DHS involvement and maternity were thetwo factors most strongly associated with dropout. 19

Figure 9.FOUR-YEAR GRADUATION RATESBY LEVEL OF DHS INVOLVEMENT

Figure 10.FOUR-YEAR GRADUATION RATES FOR ADOLESCENT MOTHERS AND NON-MOTHERS


COHORT YEAR

2008-092007-082006-072005-062004-052003-042002-030

10%20%30%40%50%60%70%80%90%

100%

No Prior DHSOther Types of DHS

Foster CareJuvenile Justice

0

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Non-Mothers Adolescent Mothers

2008-092007-082006-072005-062004-052003-042002-03


COHORT YEAR


14/24


P E R C E N T P R O B A B I L I T Y D R O P O U T

COHORT YEAR

No DHS

Other Types of DHS

Foster Care

Juvenile Justice

0

5%

10%

15%

20%

25%

2008-092007-082006-072005-062004-052003-042002-03

Students with DHS InvolvementStudents with any type of DHS involvement had a higherprobability of dropping out than students with no DHSinvolvement (Figure 11). In recent cohorts, however,DHS-involved students’ probability of dropout decreasedas they began to catch up to their non-involved peers.Students in foster care and students with other types

of child welfare involvement (for example, in-homeprotective services) had similar probabilities of droppingout over time: 15 percent and 9 percent respectivelyin the 2002–2003 cohort and 11 percent and 7 percentrespectively in the 2008–2009 cohort. Across all cohortyears, students with juvenile justice involvement had thehighest probability of dropout: three times that of non-involved students. Yet the probability of dropping out forstudents in juvenile justice decreased over time from 21percent in the 2002–2003 cohort compared to 15 percentin the 2008–2009 cohort.

Adolescent MothersAs shown in Figure 12, the probability of dropout forfemale students who have had one birth lagged behindfemale students without births as well as male students.The probability of mothers dropping out did notdecrease substantially between the 2002–2003 cohort(13 percent) and the 2008–2009 cohort (11 percent). 20

Adolescent mothers’ were consistently more likely todrop out than male students and female students with nobirth, and that likelihood increased for females with morethan one child.

Given the improvement in graduation and dropoutrates, albeit small for some groups of students, wewere interested to see how these gains related to re-engagement of students who had dropped out. Thefollowing section describes trends in re-engagement.

Figure 11.

PROBABILITY OF DROPOUT BYLEVEL OF DHS INVOLVEMENT

Figure 12.

PROBABILITY OF DROPOUT FOR ADOLESCENT MOTHERS,MALE STUDENTS, AND FEMALE STUDENTS WITHOUT A BIRT

P R O B A B I L I T Y O F D R O P O U T

COHORT YEAR

0

5%

10%

15%

20%

25%

1 Birth No BirthsMales

2008-092007-082006-072005-062004-052003-042002-03


15/24


Of the students who dropped out of high school,an increasingly large percentage re-engaged.

RE-ENGAGEMENT IN PHILADELPHIA

KEY FINDING: The percentage of all dropouts who re-engagedwith the school sys tem rose steadily from 47 percentin the 2002–2003 cohort, where 2,721 out of 5,730dropouts re-engaged, to 54 percent in the 2008-2009 cohort, where 2,090 out of 3,868 dropouts

re-engaged.

This section provides information on youthre-engagement in Accelerated High Schools,Gateway to College, the Education OptionsProgram, GED to College, E 3 Centers, andOccupational Skills Training Programs, six keyinterventions expanded or established by ProjectU-Turn collaborative members after UnfullledPromises was released in 2006.

Table 2 on page 17 provides detailed descriptions

of the programs and services that are included inthis report.

ENROLLMENT BY PROGRAM

KEY FINDING: One in ve students across all cohorts, and over half of allre-engaging dropouts, participated in at least one ProjectU-Turn program: GED to College, Educational OptionsProgram, E 3 Centers, Accelerated High School, Gatewayto College, or Occupational Skills Training Program.

The programs vary widely in the number of studentsthey serve each year. According to participation recordsanalyzed for this study, GED to College, Gateway toCollege, and the Occupational Skills Training Programeach worked with approximately 50–150 studentsannually. E 3 Centers served several hundred studentsper year, while the Accelerated High Schools andthe Educational Options Program collectively servedbetween 1,000 and 2,000 students per year. Combined,the six programs provided about 5,000 slots to strugglingstudents during their peak years. The programs also

served students in foster care, youth returning from juvenile justice placements, and adolescent mothers;these subgroups were represented in all six programs atabout twice the rate of other students.

What were the re-engagement rates for high school dropouts over time?Of the students who dropped out of high school and re-engaged, how manyparticipated in a Project U-Turn re-engagement program?

Figure 13. PARTICIPATION IN RE-ENGAGEMENT PROGRAMS, BY COHORT

0

200

400

600

800

1000

1200

1400

2008-092007-082006-072005-062004-052003-042002-03

Accelerated HSGateway to CollegeEOPOSTPGED to CollegeE3

N U M B E R O F S T U D E N T S


16/24


Over the same period in which the re-engagementprograms were introduced or expanded, the proportionof high school dropouts who chose to re-engage steadilyincreased, from 47 percent in the 2002–2003 cohort to54 percent in the 2008–2009 cohort (Figure 14). 21 And alarger proportion of re-engaging dropouts participated

in a Project U-Turn program, rather than returning to theirneighborhood school. In later cohorts, between half andtwo-thirds of all re-engaging students were involvedwith one of the six programs, with the Accelerated HighSchools and Educational Options Program serving thehighest proportion.

Graduation Rates for Former Dropoutswho Re-Engaged While the re-engagement efforts were successfulat drawing youth back into the school system andincreasing the percentage of dropouts who re-engaged,these efforts were less successful in moving students

through to graduation. The graduation rate of dropoutswho re-engaged remained steady at approximately35 percent across the seven cohorts. Thus, while there-engagement programs successfully pulled moredropouts back in, the programs may not have providedthe students with ef fective ways of achieving a highschool diploma. Still, while the graduation rate of re-engaging dropouts did not increase over time, a largerproportion of dropouts chose to re-engage. The increasein reengagement means that there were fewer dropoutsin each cohort, which was one of the ultimate goals ofProject U-Turn.

Figure 14.

PERCENT OF DROPOUTS WHO EVER RE-ENGAGED

0

10%

20%

30%

40%

50%

60%

2008-092007-082006-072005-062004-052003-042002-03

47 % 49 % 49 % 53 % 56 % 55 % 54 %

P E R C E N T R E

- E N G A G E D

COHORT YEAR


17/24


Service orProgram Description Target Population Partners

AcceleratedHigh Schools

The Accelerated High Schools are small high schoolprograms where students can accumulate credits at an

accelerated pace and earn a diploma in less than threeyears. These programs offer teacher- and computer-based instruction.

Off-track and out-of-schoolyouth ages 15–21 who are

two or more years behind inschool with fewer than halfof the credits required forgraduation accumulated.

Operatedby providers

contracted bySDP.

Gateway toCollege

Located at the Community College of Philadelphia,Gateway to College is a dual enrollment program thatlets students simultaneously earn their high schooldiploma and college credits.

Youth ages 16–21 with 21 orfewer high school credits.Eligible students mustdemonstrate the ability toread at the 8th-grade levelor higher and completerequirements for highschool graduation before orduring the year of their 22nd

birthday.

Operatedby theCommunityCollege ofPhiladelphiathrough acontractualagreementwith SDP.

EducationalOptionsProgram

Formerly known as “Twilight Schools,” the EducationalOptions Program lets students and adults continueearning credits toward a high school diploma throughafternoon and evening classes at select SDP highschools.

Youth over age 17 withat least eight high schoolcredits, but who are notcurrently enrolled in aregular day school.

Operated bySDP.

GED toCollege

GED to College provides a pathway for out-of-schoolyouth to earn a GED while also receiving support forenrolling and persisting in college. This model focuseson connecting pre-GED programming to post-GEDsuccess in college.

Out-of-school youth ages17–21 without a secondarycredential. Eligible studentsmust test at or above the7th-grade level in readingand math.

Operatedby providerscontracted byPYN.

E3 Centers E3 Centers offer a holistic approach to preparingout-of-school youth and youth returning from juvenile justice placement to achieve long-term educ ational,career, and personal goals. Services include low-literacy supports, GED-prep classes, post-secondaryaccess and planning, and intensive work-readinessprogramming that prepares participants forunsubsidized employment. Other services include job-readiness training, subsidized internships, community-service and service-learning opportunities, and jobsearch assistance.

Youth ages 16–21 who havedropped out of schooland/or are returning from juvenile justice placement.


OccupationalSkills TrainingProgram

The Occupational Skills Training Program offersopportunities for technical-skill development intargeted industries specically for out-of-school, over-aged youth. The program helps young people improvetheir academic skills and offers wrap-around servicesthat foster success in attaining a GED, an industry-recognized credential, employment, or access to anadvanced occupational-skills training institution orother higher-education institution.

Out-of-school youth ages17–21 with or withouthigh school credentials.Eligible students must testat or above a 6th-gradereading and math level andmeet federal WorkforceInvestment Act eligibility(income and barrier)requirements. Someprograms serve youth upto age 24.


Table 2. PROGRAM DESCRIPTIONS


18/24


Although graduation rates increased, rates of enrollment intwo- and four-year post-secondary institutions failed to keep pace.

POST-SECONDARY ENROLLME NT

KEY FINDINGS: • Enrollment in post-secondary institutions for all

students marginally increased.

• Students involved in DHS and adolescent motherscontinue to lag far behind in post-secondaryenrollment, and students in juvenile justice enroll inpost-secondary education at substantially lower ratesthan all other DHS categories.

Enrollment in post-secondary education reveals adifferent story than that of gains in graduation anddecreases in dropout for the 2002–2003 through 2008–2009 cohorts. For this study, enrollment rates in post-secondary institutions were reported for all rst-timeninth graders across all seven cohorts, not just thosewho graduated from high school. Later enrollment ineither a two- or four-year institution hovers between 36and 38 percent across all cohorts. 22 Foster care–involvedyouth show positive trends, as the percent of studentsenrolling in post-secondary institutions increased byapproximately 7 percentage points from the 2002–2003to 2008–2009 cohorts. Adolescent mothers’ enrollmentin post-secondary institutions declined between the2002-2003 and 2008–2009 cohorts.

LimitationsWhile the evidence reveals some important gains in highschool graduation among SDP students, the data wereevaluated only through the end of the 2011–2012 schoolyear. Since then, signicant cuts in public educationfunding for SDP resulted in highly publicized staffreductions in neighborhood schools. The district hastaken actions in an attempt to compensate for the $304million dollar cuts in State and Federal funding. Overthe two years, the District has cut over 5,000 positions,closed 32 schools, cut its administrative costs by over50%, and negotiated givebacks from two of theirunions. At the school level, this has meant larger classsizes and severe cut backs on services such as nursingand counseling services, arts and music programming,afterschool activities, and sports programs. The gainsin re-engagement are robust for the years of this study,however, as with school stafng, budget cuts may havehad a signicant impact on dropout prevention andintervention programs. What makes the ndings ofthis report signicant is that they coincide with a timeof increased funding to SDP, which may have helped

to augment services to students. With the fundingreductions in recent years, it will be important to monitorearly warning indicators (for example, early math andreading prociency, and school absenteeism), andindicators at the high school level for evidence to discernwhether the trends reported in this study are now at risk.

Many of the re-engagement programs included in thisreport support students in obtaining a GED. However,GED data were not obtained for this study. Therefore,even if a youth obtained a GED at some point, thestudent was counted as a high school dropout in this

study. Lastly, while we could track participation in ProjectU-Turn programs and knew which students earneddiplomas, we could not link the two directly to determinewhich students or how many had earned their diplomaswhile enrolled in one of the intervention programs.As such, the credential attainment rate of differentprograms could not be measured.

How many students eventually enrolled in post-secondary institutions?

Figure 15. POST-SECONDARY ENROLLMENT

Note: Enrollment rates in two- and four-year institutions for allrst-time ninth graders in each cohort, as of April 2014.

05%

10%

15%20%25%

30%35%40%

45%

MaternityOther DHS ServicesFoster Care Juvenile JusticeAll Students

2008-092007-082006-072005-062004-052003-042002-03

P E R C E N T E N R O L L E D

COHORT YEAR


19/24


CONCLUSIONS

The data presented in this report provide evidence that graduation rates have

increased while dropout rates have decreased for Philadelphia public school students

over time, yet enrollment rates in two-year and four-year post-secondary institutions

have not kept pace. Graduation rates for higher-risk populations, such as students

with DHS involvement and female students who gave birth prior to and/or during high

school, remain below the overall cohort graduation rate. In addition, male studentscontinue to graduate at lower rates than female students. The data also demonstrate

that a higher percentage of high school dropouts are re-engaging in Project U-Turn-

afliated programs. Despite this, the graduation rate for dropouts who re-engaged

remained fairly steady.

This report shows that Philadelphia public schools are graduating considerably more

students than in the past. More can be done, however, to ensure that those being left

behind receive the support needed before they drop out. Keeping students on the

path to on-time graduation is key to reducing the number of high school dropouts. Thendings in this report provide the City of Philadelphia, SDP, and other partners and

stakeholders with the evidence for continued systemic change to ensure a better future

for Philadelphia’s students. The ndings also afrm the benet of data sharing between

SDP and DHS to track the educational outcomes of students in child welfare, who were

found to be at the highest risk. Data sharing among public systems enables alignment

of resources to meet the needs of the varied student population in Philadelphia public

schools. Given limited public resources, it is important to provide support for our most

at-risk populations. With the high proportion of at-risk youth enrolled in SDP, aligning

support and services to engage students around school persistence and academicachievement will be critical to continued improvement.


20/24


REFERENCESAllison, P. D. (1982). Discrete-time methods for theanalysis of event histories. Sociological Methodology , 13,61–98.

Aron, L., & Loprest, P. (2012). Disability and the educationsystem. The Future of Children , 22 (1), 97-122.

Aud, S., Hussar, W., Kena, G., Bianco, K., Frohlich, L.,Kemp, J., & Tahan, K. (2011). The condition of education2011. NCES 2011-033. Washington, DC: National Centerfor Education Statistics.

Balfanz, R., Byrnes, V., & Fox, J. (2013). Sent home andput off-track: The antecedents, disproportionalities, andconsequences of being suspended in the ninth grade.The Center for Civil Rights Remedies . Retrieved from:http://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-remedies/school-to-prison-folder/state-reports/sent-home-and-put-off-track-the-antecedents-disproportionalities-and-consequences-of-being-suspended-in-the-ninth-grade/?searchterm=senthome and put off track

Basch, C. E. (2011). Teen pregnancy and the achievementgap among urban minority youth. Journal of SchoolHealth , 81(10), 614–618.

Chang, H. N., & Romero, M. (2008). Present, engaged,and accounted for: The critical importance of addressingchronic absence in the early grades. Report. NationalCenter for Children in Poverty.

Gasper, J., DeLuca, S., & Estacion, A. (2012). Switchingschools: Revisiting the relationship between schoolmobility and high school dropout. American EducationalResearch Journal , 49 (3), 487–519.

Hirscheld, P. (2009). Another way out: The impact of juvenile arrest s on high school dropout. Sociology ofEducation , 82 (4), 368–393.

Hjalmarsson, R. (2007). Criminal justice involvement andhigh school completion. Journal of Urban Economics ,63 (2), 613–630.

H.R. 6893 (110th Congress): Fostering connections tosuccess and increasing adoptions act of 2008. In: PL 110-351, Section 204, (2008).

Hwang, S., Grifs, H., Song, L. & Rubin, D. (2014).Supporting the needs of students involved with thechild welfare or juvenile justice system in the SchoolDistrict of Philadelphia. Philadelphia, PA: PolicyLab at TheChildren’s Hospital of Philadelphia.

Kim, J. (2011). Relationships among and between ELL status, demographic characterist ics, enrollment history,and school persistence (CRESST Report 810). LosAngeles, CA: University of California, National Center forResearch on Evaluation, Standards, and Student Testing(CRESST).

Neild, R. C., & Balfanz, R. (2006). Unfullled promise: Thedimensions and characteristics of Philadelphia’s dropoutcrisis, 2000–2005. Philadelphia Youth Network.

Rumberger, R., & Lim, S. A. (2008). Why students dropout of school: A review of 25 years of research.Retrievedfrom: http://cdrp.ucsb.edu/dropouts/pubs_reports.htm.

Schoeneberger, J. A. (2012). Longitudinal attendancepatterns: Developing high school dropouts. The ClearingHouse: A Journal of Educational Strategies, Issues andIdeas , 85 (1), 7–14.

Sheng Z., Sheng, Y., & Christine J. Anderson (2011)Dropping out of school among ELL students: Implicationsto schools and teacher education. The Clearing House: A

Journal of Educational Strategies, Issues and Ideas , 84 (3),98–103.

Stetser, M., and Stillwell, R. (2014). Public high schoolfour-year on-time graduation rates and event dropoutrates: School years 2010–11 and 2011–12 . First Look (NCES2014-391). U.S. Department of Education. Washington,DC: National Center for Education Statistics.


21/24


22/24


Variables

SDP student data includes gender, race/ethnicity,grade level, special education status, limited Englishprociency, absences, suspension, grade promotion,and nal enrollment status. Final enrollment status

consists of four categories: dropped out, graduated,continuing, or transfer. Students continuing are thosethat remain enrolled and those who transfer to anyschool outside of SDP.

Students were also identied if enrolled in theEducational Options Program, Accelerated HighSchools, or Gateway to College program offeredthrough SDP. Through SDP, the National StudentClearinghouse data was provided for post-secondaryenrollment, which identied any student enrolling in atwo and/or four year institution.

Students involved with both the child welfare and juvenile justice systems (that is, “crossover youth”) werehierarchically classied so categories were mutuallyexclusive and students were not double-countedin this analysis. Each student was classied as: 1) juvenile justice 2) foster care 3) other DHS 4) no DHSinvolvement. This was considered so that students athighest risk would be labeled as such, and studentswho were only served by DHS were placed in a lowerrisk category. Thus, a student who received fostercare and juvenile justice services at various points in

time would be included only in the “juvenile justice”category. A student who received both foster care andother services, such as prevention services, would beincluded only in the “foster care” category.

Methods

Dropout by Grade: Using the 2011–2012 school yearsnapshot of all students in high school, the dropoutrate for each grade was calculated by selecting anystudent with a nal status of “dropout” at the end of the

school year and dividing by the total number of studentsstarting each grade at the beginning of that school year.

Graduation and Dropout calculation: Four-year dropoutand graduation calculations were based on the nalstatus of the student provided by SDP, which includeddropped out, graduated, continued, or transferred out ofSDP. These were calculated for each cohort. 24 GEDs anddiplomas obtained outside the district were not includedin the calculation of graduation or dropout. Rates donot include students who began in a charter school inninth grade or students transferring outside of SDP after

starting ninth grade. Students transferring into SDPschools were not included in the calculations.

Probability of dropout: Discrete time hazard modelswere used to examine the effect of a student’s changein high-risk status (for example, receiving child welfareservices or having a child) on the risk of dropout. Discretehazard models account for time and other variables, suchas grade and race/ethnicity, and model the “time-to”an event—in this case, time to drop out. 25 Each studentbegan an observed risk period upon entering ninthgrade. When a student dropped out, the risk period

ended. Every student who graduated was right-censoredso as to not incorrectly be counted as a dropout. Right-censoring allowed for graduating students to leave thestatistical model as a graduate or continuing student,rather than counting as a dropout. Robust standarderrors were used to account for the correlation due toclustering of students within schools. These models werecalculated separately for each cohort and adjusted forother student characteristics over time, including race/ethnicity, gender, DHS involvement, special educationstatus, limited English prociency status, and school year.


23/24


ENDNOTES1 Rumberger & Lim, 2008

2 Dianda, 2008

3 Neild & Balfanz, 2006

4 Description of the data is available inAppendix 2.

5 See http://www.projectuturn.net/research.php for research and reportsgenerated since 2006.

6 The percentage of students in eachgrade transferring to schools outsideSDP ranged between 1 and 4 percent,with the lowest percentage in twelfthgrade.

7 These analyses do not include ninth-

grade students star ting in charter,private, or parochial schools. Studentswho started in an SDP school in ninthgrade and subsequently transferred toa charter school are included in theseanalyses. A description of high schooltypes is available in Appendix 1.


9 Stetser & Stillwell, 2014


11 SDP student enrollment numbers also

declined during this period.

12 Schoeneberger, 2012

13 Chang & Romero, 2008

14 Balfanz, Byrnes, & Fox, 2013

15 DHS categories are hierarchical andmutually exclusive.

16 Hwang, Grifs, Song, & Rubin, 2014

17 Aud, et al, 2010; Rumberger & Lim,2013; Kim, 2011; Aron & Loprest, 2012;Sheng, Sheng & Anderson, 2011

18 Gasper, DeLuca, & Estacion, 2011;Basch, 2011; Hirscheld, 2009;Hjalmarsson, 2007

19 Other factors examined includegender, race/ethnicity, grade level,receipt of special education services,limited English prociency, and schooltype.

20 Students involved with DHS and teenmothers often stay in high schoollonger than other students. Therefore,lower dropout probabilities may bean artifact of including students s tillenrolled in high school who may dropout after their fourth year in highschool.

21 In terms of demographicbackgrounds, the dropouts who choseto re-engage were no different thanthose other dropouts who remainedout of school. The percentages ofstudents by gender, ethnicity, limited

English prociency, special education,foster care placement, juvenile justice,and maternity were the same for re-engaging students and dropouts whonever returned.

22 Post-secondary enrollment recordswere available through spring 2014,giving the 2002–2003 cohort ofninth graders up to eight years af teranticipated high school graduation toenroll in a post-secondary institution,compared to only two years forthe 2008–2009 cohort. Thus, post-secondary enrollment rates formore recent cohorts could still riseslightly as some students enroll inpost-secondary schooling for therst time more than two years aftertheir expected time of high-schoolgraduation.

23 Derived from listings on the SchoolDistrict of Philadelphia, Ofce ofStudent Enrollment and Placementwebsite http://webgui.phila.k12.pa.us/ofces/s/student-placement/school-selection-process and from Hwang,Grifs, Song, & Rubin, 2014, p. 8

24 Rates are based on August 2014 dataand reect an individual’s status fouryears after entering high school.

25 Allison, 1982.

This project would not be possible without the support of many civic leaders whose concern for the youthof Philadelphia led to this research.

This work was supported by the generosity of the William Penn Foundation. We thank the School District ofPhiladelphia and the Department of Human Services for their guidance and collaboration as we navigatedthe data to create this report. We express our sincere thanks to Sophia Hwang, doctoral s tudent, New YorkUniversity, LiHai Song, Biostatistician, and Meredith Matone, Research Scientist, at the PolicyLab at TheChildren’s Hospital of Philadelphia. Finally, we thank members of the Project U-Turn steering committee,and staff f rom the Philadelphia Youth Network for reviewing this report and providing insightful feedback.

The opinions expressed in this report are those of the author(s) and do not necessarily reect the views ofthe William Penn Foundation.


24/24

For more information aboutProject U-Turn, please contact:

267-502-3800www.projectUturn.net

promise worth keeping

Documents