Technical ReportPDF Available

Transportation Equity in a School Choice Program

Authors:

Abstract

This report presents findings from a multi-method analysis of school transportation in the Greater Hartford School Choice program. The study was commissioned by the Connecticut Department of Education’s Regional School Choice Office (RSCO) per the Sheff Comprehensive Choice Plan (CCP).1 The CCP calls for a reassessment of the transportation program “to ensure that present practices do not inadvertently create significant disincentives for participation in Choice programs" (p. 36). The report presents four sets of studies, followed by our recommendations based on the findings.
RSCO Transportation Study Report
Submitted by
Casey D. Cobb
Charles Wentzell
Kelly Farrell
University of Connecticut
December 30, 2023
2
Executive Summary
This report presents findings from a multi-method analysis of school transportation in the Greater
Hartford School Choice program. The study was commissioned by the Connecticut Department of
Education’s Regional School Choice Office (RSCO) per the Sheff Comprehensive Choice Plan (CCP).
1
The
CCP calls for a reassessment of the transportation program “to ensure that present practices do not
inadvertently create significant disincentives for participation in Choice programs" (p. 36). The report
presents four sets of studies, followed by our recommendations based on the findings.
Study 1
The first is a qualitative study of interviews with parents about their family’s experience with
transportation to and from school. Our case study design allowed for exploration of the conditions,
processes, and beliefs that shape families’ experiences with transportation which, in turn, informs their
decisions to enroll their children in a magnet or Open Choice school. We identified five emergent
themes from the analysis of interview data. They included: bus stops, parent adaptations and personal
transportation, after school extracurriculars, communication, and sacrifices for school opportunities.
Of the 44 parents interviewed, 71 percent expressed a concern with transportation. Five parents in our
sample removed their child from the Choice program at least in part due to a transportation issue. By far
the most common complaint involved bus stops. Fifty-nine (59) percent of parents referenced problems
with the bus stop, particularly central stops. Complaints included the distance to the stop from home,
the nature of the commute to the stop, and contextual aspects of the stop location itself. Bus
transportation was significantly less accessible for suburban magnet parents who lacked personal
transportation. Two-thirds (68 percent) of parents we interviewed drove their children either directly to
school or to the bus stop each day. A handful of parents mentioned that they were disappointed that
extracurricular activities at school were inaccessible to their children due to lack of transportation. Other
families who were offered transportation for after-school activities expressed their frustration,
mentioning that the bus was often unreliable. Approximately half of the parents (48 percent) mentioned
that the communication system between the transportation service and parents was unreliable. Even
though 71 percent of parents communicated that they had a concern with some aspect of school
transportation, many felt the education their children were receiving outweighed the challenges they
were facing and continue to send their child/ren to Open Choice or magnet schools.
Study 2
The second study is a quantitative analysis of RSCO Transportation bus complaint logs from the 2022-23
school year. We descriptively analyzed complaints by ticket source, ticket date, and complaint type.
More than three-quarters (77.4 percent) of the complaints were submitted by phone, followed by the
1
Permanent Injunction, Sheff v. O’Neill, Superior Court, judicial district of Hartford, Docket No. HHD-CV17-
S040566S (January 27, 2022). Retrieved from https://ctschoolfinance.org/resource-assets/Sheff-Permanent-
Injunction.pdf.
3
online web form (15.5 percent), and email (4.0 percent). Unsurprisingly, most complaints were issued
early in the school year, a time when buses and families are adjusting to the new routine. As the year
progressed, complaints tapered off. Most complaints fell under the categories of “bus/vehicle driver,”
“late or no notification of delays,” and a category labeled as “other.” However, we observed some
misalignment between the assigned category and the actual complaint. Part of the issue was that some
of the categories overlapped with one another (e.g., “central stops” and “stop location”). Another could
be that certain complaints implicated more than one reason code but only one could be selected. "Stop
location” and “central stops” were not prominent concerns, at least as recorded in the complaint
database. Improving upon the complaint system could help better inform school transportation officials.
Study 3
The third study is a quantitative analysis of student travel times to bus stops and schools for 10,186
students who had received first-round offers to either a magnet (n=9,421) or Open Choice school (765)
in 2022-23. We estimated bus ride times using bus route schedules and geospatial techniques, and
reported these by student resident group (i.e., Hartford or Suburban), choice program, and sending
district. We also disaggregated data by distinct ride time thresholds. Ride times over 30 minutes were
considered a “long” ride and any ride over 60 minutes were deemed “very long.” Among Open Choice
students, 88.1 percent had a long expected bus ride, and 18.6 percent had a very long expected ride,
compared to 72.3 percent and 11.0 percent, respectively, for magnet school students. Relative to
Hartford-resident students, suburban magnet students generally have longer bus rides with a median of
39 minutes compared to 36 minutes for Hartford-resident students.
Based on “bus stops” as a prominent theme that emerged from parent interviews, we analyzed the
theoretical distance students would have to travel to reach their bus stops from their homes. This also
allowed us to evaluate theoretical walkability for students to stops. The results showed a wide range of
travel times for students, with the widest difference being between Hartford students (in both magnet
schools and Open Choice) and suburban magnet students. Hartford-resident Open Choice students had
the lowest median walking distance to their stop (2 minutes). Hartford-resident magnet students had a
median walking travel of 4 minutes, with a slightly higher range of distances. Suburban students in
magnet schools had a median walking travel time from home to their bus stop of 35 minutes. More than
half (54.5 percent) of suburban students in magnet schools had a walking travel time of 30 minutes or
more, indicating that for most of these students walking to their bus stop would be difficult, if not
unfeasible.
Study 4
Finally, the fourth section is a set of quantitative analyses examining the degree to which travel distance
to school was related to parent decisions to accept or decline a lottery placement offer. As in study #3,
we used the lottery data as the basis for the analysis. The lottery data contained placement offer
“accept/decline” fields, which we linked to (theoretical) estimated travel times from student home
address. We found a moderate inverse relationship (r= -.453) between estimated median driving time
and magnet seat acceptance rate per sending town. This suggests that, generally, when travel time
increases, acceptance rates decrease but not in a perfect linear fashion. While suburban and Hartford
students who accepted magnet placement offers had an overall lower median estimated travel time
than students who actively declined17 (15 minutes vs. 17 minutes, respectively), this difference was
relatively small, especially in comparison to the between-school differences and between sending
districts for magnet students.
4
In our logistic regression model, we five variables to predict parent acceptances. We found that the
strongest predictor of accepting a lottery magnet seat was the school ranking variable in the lottery
application. For every standard deviation unit increase in school preference score, the odds of accepting
a magnet seat increase by 49.7 percent. The next strongest influence on accepting was being a Hartford
resident, which increased the odds of accepting a magnet seat by 41.5 percent. The remaining variables
exhibited a negative influence on parents’ decision to accept a seat, with being a RI student having the
strongest relationship. Notably, driving time to school had virtually no effect on parent decision making,
after taking into account the remaining variables in the model.
Summary of Suggestions Based on Findings
We offer recommendations for policy and practice and other ideas for consideration to improve the
school transportation experience for students and parents. Beyond practical consequences, we consider
the implications for equity and fairness for families participating in Choice.
Improve conditions for getting students to and from the bus stop.
Ensure all stop locations are safe.
Recalibrate the complaint type categories in the RSCO online complaint form.
Make the online complaint form more prominent and accessible on the website.
Recalibrate bus notification system to improve efficiency.
Ensure all families have access to bus notification mechanisms.
Involve families and students in developing transportation policies.
Look into other transportation models.
Communicate transportation options to prospective Choice parents.
Consider walking chaperones for younger students.
Explore offering free discounted or public transit passes for age-appropriate students.
Investigate further regionalizing school choice.
5
Introduction
Proponents of school choice argue it levels the playing field for students who otherwise would not have
access to quality schools (Chubb & Moe, 1990). Others find school choice to be more of an illusion,
working for some families and not others (Orfield & Frankenberg, 2013; Pattillo, 2015). Research has
shown how race and class can either enhance or constrain choices available to families (Phillips et al.,
2012). Racially minoritized and low-income students are disproportionately impacted by the constraints
(Sattin-Bajaj & Roda, 2020). Even among those who participate in school choice programs, sacrifices are
often made. For students, these may include longer school commutes, less sleep, or loss of connection
with neighborhood friends. For parents, getting their child to and from school may take time away from
work or require expending scarce resources on a car or gas. For school choice to realize its promise,
policies must preempt or minimize such potential barriers.
Despite being an integral component of school choice, student transportation is often an overlooked or
at best underexamined feature. One of the major factors affecting parent participation and satisfaction
with school choice is ensuring their child safe travel to and from school. Attending a school outside the
local district typically coincides with an increased transportation burden for students and their families.
Some families participating in school choice drive their children to school themselves, provided they
have the means to do so. Families without this option, or who prefer not to drive their child, rely on bus
transportation to a choice school or may not participate in choice altogether. Even those who use bus
services may feel compelled to drive their child to a centralized stop. All of this is to say that school
choice transportation raises questions of practicality and equity.
Research on mobility justice suggests that unequal access to transportationand, sometimes, a
need to travel long distances to reach desirable schoolscan make transportation a key factor
in shaping equitable access to schools in choice-oriented settings (Bierbaum et al., 2021) (as
cited by Valant & Lincove, 2023, p. 535)
For voluntary choice programs designed to desegregate schools, access to safe, affordable, and efficient
bus transportation is paramount to achieving the desired outcome. This report presents findings from a
mixed-method investigation on transportation in the Greater Hartford School Choice program. The
program is overseen by the Connecticut State Department of Education’s Regional School Choice Office
(RSCO). Our study focused on the perspectives of families participating in the program and how they and
their children experienced travel to school. In addition to interviews with families, we drew on several
other sources, including transportation complaint logs collected by the bus contractor, busing schedules,
school choice lottery data, as well as reviews of RSCO transportation documents and the relevant
research literature. The study assists the state's obligation to meet Commitment #33 of the Sheff
Comprehensive Choice Plan,
2
which calls for a reassessment of the transportation program to ensure
2
Permanent Injunction, Sheff v. O’Neill, Superior Court, judicial district of Hartford, Docket No. HHD-CV17-
S040566S (January 27, 2022). Retrieved from https://ctschoolfinance.org/resource-assets/Sheff-Permanent-
Injunction.pdf.
6
that present practices do not inadvertently create significant disincentives for participation in Choice
programs" (p. 36).
The report is organized as follows. We begin by summarizing the literature on school choice
transportation, followed by a brief overview of the RSCO Greater Hartford School Choice program and
its associated transportation services. Next, we present analytic approaches and findings from our multi-
method design; they are organized by data sources including parent interviews, bus complaint logs,
student location and bus ride data, and student magnet lottery data. We end by offering policy and
practice recommendations and other ideas for consideration.
Research on School Choice Transportation
Distance to school is an important consideration influencing family decisions regarding school choice,
regardless of grade level or the type of school (Burgess et al., 2015; Hastings et al., 2005). All other
things equal, long commutes are generally worse than short ones, given earlier wake-up times for some
students, the extra stressors for students enduring long rides, less time for homework or extracurricular
activities, and likelihood for increased absenteeism.
3
Other issues may factor into how students and
families experience bus transportation, including the reliability and timeliness of pickups and drop-offs,
bus stop locations, the distance between home and bus stop, the number of stops or transfers, the
availability of after school late buses, and of course the experience on the bus itself as it relates to issues
of comfort and safety.
Influence of Transportation and Distance on School Choice Participation
Increasing evidence suggests that transportation significantly influences school enrollment choices.
Teske et al. (2009) surveyed 600 parents of K-12 children from lower-income backgrounds in Denver and
Washington DC, which included both school choosers and parents sending their children to
neighborhood assigned schools. Among the respondents, 38 percent highlighted transportation as an
issue influencing their school choices. Notably, 27 percent identified a preferred school that they did not
pursue due to transportation-related concerns. In Denver, families with children attending their
neighborhood assigned schools prioritized "location/convenience" nearly five times the rate of those
who chose schools outside their locally assigned one (44 percent compared to 9 percent) (Teske et al.,
2009, p. 16).
In their study of kindergarten students in New York City, Trajkovski et al. (2021) found that having
reliable school transportation or living close to the school improves the likelihood of parents
participating in school choice. Cordes and Schwartz (2018) explored the link between transportation and
school choice among elementary students in New York City, finding that bus riders show a higher
tendency to attend a choice school over their zoned school. Stein et al. (2020) and Blagg et al. (2018)
found that extended or taxing school commutes prompted certain students to transfer to schools nearer
their homes. Yettick (2016) discovered that location, rather than school ratings, emerged as the primary
factor for parents' initial selection of schools, especially among low-income parents.
3
There may be some positive aspects to longer bus rides (e.g., families who do not have access to before or after
school daycare); however, even these potential benefits have their own tradeoffs.
7
Stein et al. (2021) analyzed public transit commute times in Baltimore and found that high school
students with commutes exceeding an hour were approximately three times more inclined to switch
schools compared to those with commutes under 10 minutes. Further, they reported:
[S]tudents who do change schools, on average, attend new schools that are closer to home but
less likely to have been ranked highly [in the top 5] in their initial choice application. It is
possible that these associations could be even stronger if we had measures of actual commuting
behavior and not just travel estimates (p. 142).
A survey among parents of K-6 students in the city of St. Paul, MN, and adjacent suburban Roseville Area
Schools indicated that school choice resulted in longer school commute distances and decreased levels
of walking and bicycling to school (Wilson et al., 2010). At the time, St. Paul was home to 34 intra-district
magnet schools while Roseville had one intra-district magnet school. The study also found that magnet
school students typically had a longer commute distance, with a median travel distance of 2.7 miles for
magnet compared to 1 mile for neighborhood schools.
Lincove and Valant (2018) examined commute times by school bus, car, and public transit among a
sample of 17 charter and district-run schools in New Orleans (Lincove & Valant, 2018). Car travel was by
far the shortest commute time (median time of 14 minutes), followed by school bus (35 minutes) and
public transit (minutes); car travel time also exhibited the least variability. In their analysis of morning
bus ride time for 120,000 New York City elementary students, Cordes et al. (2022) found that the typical
bus ride lasted around 21 minutes, with most students having commutes of less than 30 minutes. Only
6.1 percent of students had long bus rides of between 45 to 60 minutes, while rides exceeding one hour
affected 3.3 percent of bus riders. However, they also found an unequal impact on students of color
who participated in school choice:
Students with very long bus rides are disproportionately Black and almost exclusively attend
district choice or charter schools. Commute times negatively impact both attendance and
chronic absenteeism, particularly among students in district choice schools, for whom long and
very long commutes decrease attendance by 0.330 and 0.625 percentage points and increase
the probability of chronic absenteeism by 1.8 and 3.2 percentage points, respectively (Cordes et
al., 2022, p. 690).
Lenhoff et al. (2023) researched student transportation in choice-rich Detroit public schools. Slightly
more than half (53 percent) of the students were eligible for school-sponsored transportation (i.e.,
school bus or public bus passes). They found that roughly half of the students who were eligible to ride
the bus never did so.
There is also some evidence that bus access and use is related to student race and socioeconomic status
(Weinstein et al., 2022). Research on school bus transit in New York City reported that Black students
were less likely to use school buses compared to White students, even when both groups resided at
similar distances from school (Weinstein et al., 2022). Additionally, other studies have indicated that
Black students tend to commute to school by car, while students from higher socioeconomic
backgrounds are more prone to using the bus (Rhoulac, 2005).
8
Distance and Attendance
Some evidence suggests choice students who take the bus exhibit higher attendance and are less likely
to be chronically absent [once enrolled] (Cordes et al., 2019; Gottfried, 2017). Gottfried (2017) found
that reliable school transportation was associated with higher academic and non-academic outcomes,
such as attendance. Cordes et al. (2022) reported that longer commute times were related to lower
attendance and higher chronic absenteeism in New York City district choice schools. Other research
demonstrated that long commutes were associated with reduced sleep and exercise for students
(Voulgaris et al., 2019).
Equity Considerations
School desegregation programs that operate under voluntary school choice are intended to promote
educational equity. Research shows that historically marginalized groups participating in school choice
generally face longer commutes (BurdickWill, 2017; Corcoran, 2018; Cowen et al., 2018; He & Giuliano,
2018; Scott & Marshall, 2019; Stein et al., 2017). If under-resourced families are disparately burdened by
transportation issues or, worse yet, they avoid participating altogether due to difficulties with
transportation, equity remains elusive. Transportation equity research can help inform the degree to
which educational access and transportation equity are at odds (Bierbaum et al., 2021). For some
students, commute times can be long because choice schools are often located lengthy distances from
student residences. Returning home from school can also be complicated by limited late bus options for
students participating in after-school activities. Relieving burdens on choice students leads to more fair
and reasonable transportation experiences and could invite increased participation in choice.
Additionally, household structure and social networks are influential in shaping how students commute
to school (Bierbaum et al., 2021). Single-parent households or those with two working parents tend
toward car usage due to its flexibility in managing complex travel logistics (Makarewicz, 2013; Mandic et
al., 2017). While women often serve as primary caregivers and frequently accompany children to school,
regardless of transportation mode, challenges in work schedules and safety perceptions hinder walking
and cycling to school (He, 2013; Lidbe et al., 2020).
The Greater Hartford School Choice Program
School choice is designed to improve educational outcomes and provide educational equity for students
in highly segregated schools. In Hartford, voluntary public school choice is offered through interdistrict
magnet schools, vocational schools, charter schools, and an interdistrict student transfer program. This
study focuses on the interdistrict magnet schools and Open Choice program in the Greater Hartford
region, two programs overseen by RSCO. Both programs are designed to desegregate racially, ethnically,
and economically stratified schools as part of the longstanding Sheff v. O’Neill (1996) settlement
agreement.
Greater Hartford is home to 43 interdistrict magnet schools ranging in grade levels and academic or
curricular foci. Most magnet schools are operated by either Hartford Public Schools (Hartford Host
Magnets) or the Capitol Region Education Council (CREC Magnets). CREC oversees the Hartford-area
Open Choice program
4
, an urban-suburban student transfer program that encourages students from
4
https://schoolstatefinance.org/resource-assets/Connecticuts-Open-Choice-Program.pdf
9
Hartford to attend one of several suburban
5
school options within their geographic zone, and suburban
students to attend a school in Hartford. In 2022-23, more than 20,000 Greater Hartford students
attended the two largest interdistrict choice programs -- magnets and Open Choice. Aggregate magnet
school enrollment is about seven times that of Open Choice. Seats are made available by school districts
on a voluntary basis.
School choice applications are administered through an annual lottery managed by RSCO. Interested
families submit applications indicating their school preferences and can rank up to five schools per
magnet and Open Choice programs. Student placement is dictated by several factors, including
preferences for applicants with siblings enrolled in a choice school. In addition, a primary placement
factor involves a measure of family socioeconomic status (SES). The current guidelines for placement call
for new cohorts to be assigned to schools such that they enroll no more than 60percent students from
SES Tier A (lower income students) and no lower than 30 percent students from SES Tier C (higher
income students). The settlement stipulation includes other requirements, such as meeting Hartford-
resident demand targets and ensuring enrollment in an interdistrict magnet does not exceed 75 percent
from any single school district, including Hartford. While the RSCO lottery offers placements to schools,
families may decline offers and do so at reasonably high rates; in the 2019-2020 lottery, Hartford-
resident students declined a first-choice magnet school offer 43.3 percent of the time while Suburban
students declined at the modestly higher rate of 56.7 percent (Cobb & Connery, 2021).
Influence of Transportation on RSCO Participation
Families participating in voluntary school choice programs are particularly reliant on safe and efficient
transportation services. Busing in the Greater Hartford School Choice Program is highly complex given
the geographic spread of the region, the number of students requiring transportation, and the number
of schools involved. The RSCO transportation zone constitutes 43 municipalities in metro Hartford. Any
one magnet school may enroll students from dozens of municipalities in and around Hartford. Some
students attending interdistrict choice schools make multiple transfers to get to their destinations, as
centralized stops may be necessary to efficiently transport students. Last year, RSCO transported over
14,000 students to 180 schools across metro Hartford.
6
A previous RSCO analysis of why families declined a lottery seat suggested transportation, although not
the main reason, influenced their decisions (Cobb & Connery, 2021). The study found that 17.5 percent
of parents who actively declined their first-choice (i.e., top-ranked choice indicated by applicant) magnet
school placement offer cited transportation or travel as a concern. Of these decliners, 10.9 percent
referenced a transportation issue
7
and another 6.6 percent indicated the school or district was too far
5
An important note about our use of the term “suburban” throughout this report. We use it to reference suburban
municipalities and districts, even though there is variation in their geography and degree of “urbanicity.” There are
ostensibly “inner ring” municipalities around Hartford and an extended “outer ring” (some considered “exurbs”).
6
Source: https://www.crec.org/transportation/rsco.php
7
Transportation: not available (3.9%), pick-up/drop-off times inconvenient (3.9%), not available until after school
starts (1.6%), centralized stops are too far away (1.0%), bus stops for siblings in choice in different locations (.5%).
Source: Cobb & Connery, 2021, Table 15a, p. 22.
10
away. A similar study examined reasons why students voluntarily left a RSCO magnet school (Cobb et al.,
2021). In that analysis, 45 parents (representing 57 children who left) were asked their reason(s) for
leaving. Transportation ranked fourth among reasons, cited by 22 percent of parents.
The two prior studies suggest that transportation influences participation in the Greater Hartford
Regional School Choice Program. Of course, there are limitations to such inferences about how large a
role transportation plays in parents’ participation in RSCO. For instance, we did not know the degree to
which transportation affected decisions for families who did not apply to the lottery.
We focused on conducting a qualitative analysis of parental experiences with transportation through
interviews as detailed in this report. From this, we additionally worked to support and contextualize
themes that emerged in interviews by using secondary datasets concerning both parental complaints
and student transit details. We present the findings of these analyses followed by recommendations to
help mitigate possible negative impacts that transportation related issues may have on school choice
decisions, as well as to build on existing strengths to improve student transportation going forward.
11
Study 1: Parent Experiences with School
Transportation
For this qualitative analysis, we employed a case study design that allowed us to investigate how
parents
8
of students in interdistrict choice experienced transportation to and from school as
embedded in a real-world social context (Yin, 2018). Our case study design allowed for exploration of
the conditions, processes, and beliefs that shape families’ experiences with transportation which, in
turn, shapes their decisions to enroll and continue to enroll in their child/ren to Open Choice and/or
magnet schools. Primary data were collected via parent interviews.
This analysis was guided by two overarching research questions:
1. How do parents of children enrolled in an interdistrict magnet or Open Choice school
experience school transportation?
2. How and to what degree do their experiences influence their participation in Choice
programs?
Analytic Approach
RSCO provided a database containing student information on the interdistrict magnet or Open Choice
school that they attended alongside parent contact information. This information included students who
attended magnet or Open Choice schools at any time between the 2017-18 and the 2022-23 school
years. Using stratified random sampling, we selected students across various municipalities, schools, and
grade levels.
We employed rolling random selection contacting 25 to 50 parents at a time to obtain a sample size
sufficient to achieve data saturation (Glazer & Strauss, 2017). This occurred over a span of five months,
from June to November 2023. In total, over 500 parents were contacted via phone through text
message, or email, asking for their participation in a short interview about their transportation
experiences. As an incentive, we offered a $25 gift card for their time. We were mindful of our sample’s
representation as we scheduled participants. Thus, midway through the project, we adjusted our
sampling frame to increase participation from Hartford residents, who were underrepresented in our
sample. Response rates were low, generating 44 participants, but not altogether inconsistent with other
forms of survey research.
We conducted virtual or phone interviews with all 44 parents, who collectively represented 71 children
who were current or former magnet or Open Choice students. We asked parents about their
8
For brevity, throughout this report we use the term “parents” to more broadly refer to parents, caretakers, or
guardians.
12
experiences with school transportation and to what extent transportation affected their decision-
making process regarding initial and continued participation in choice programs. Of the 44 parents
interviewed, 18 were Hartford residents and 26 were suburban parents (Table 1). Ten (10) of the
Hartford parents sent their children to Open Choice, 5 sent their children to a magnet school, 2 sent
their children (siblings) to both a magnet and Open Choice school. Among the suburban parents, most
(84.6 percent) were enrolled in a magnet. Five parents had withdrawn their child/ren from the Choice
program.
Table 1. Parent Interview Sample (n=44)
Residence
Magnet
Magnet & OC
No Longer in
Choice
Total
N
%
N
%
N
%
N
%
N
Hartford
5
18.5%
10
100.0%
2
100.0%
1
20.0%
18
Suburban
22
81.5%
0
0.0%
0
0.0%
4
80.0%
26
Total
27
100.0%
10
100.0%
2
100.0%
5
100.0%
44
Data were collected via semi-structured interviews (Seidman, 2013). Each participant was interviewed
once for approximately 15-20 minutes. See Appendix A for the interview protocol. All interviews were
audio recorded and transcribed, with all identifiable information removed.
Transcript data were analyzed thematically via a process that included multiple rounds of inductive
coding. Following systematic procedures, we moved from narrow units of analysis (e.g., significant
statements) to broader units (e.g., meaning units) (Creswell & Poth, 2018). The goal was to understand
and describe how families experienced school transportation.
We engaged in an initial phase of inductive coding to follow themes surfacing from the data. We
organized this first phase of coding in a codebook, including for each code a definition, an example,
notes, and a frequency count. Following this first phase of coding, we conducted a second phase of
focused coding to find prominent emergent themes pertinent to the research questions (Miles et al.,
2014). During this process we refined the codebook and wrote analytic memos. This iterative process
also included frequent returns to the data set to ensure a closer and more accurate interpretation of the
data. After coding all data, we created a variety of matrices to help develop interpretations and check
for disconfirming evidence (Miles et al., 2014). Finally, to limit research bias in analysis and
interpretation of data, we conducted peer debriefing and engaged in reflexivity (Creswell & Poth, 2018).
We accomplished this by cross-checking codes and member checking of emergent themes to avoid bias
and increase trustworthiness and dependability.
Findings
Through our coding process and subsequent analysis, we identified five distinct themes that emerged
from the interview data. These included: bus stops, parent adaptations and personal transportation,
after school extracurriculars, communication, and sacrifices for school opportunities. We expound on
each below.
13
Bus Stops
Of the 44 parents interviewed, 59 percent referenced problems with the bus stop
9
as an issue with
transit for their students, particularly with regard to central stops. Complaints included the distance to
the stop from home, the nature of the commute to the stop, and contextual aspects of the stop location
itself.
Distance to the stop was a major concern for parents in our sample, who often noted that that the
location of the bus stop was too far from their home for their children to walk. This was particularly
common for parents of suburban students attending magnet schools. One parent said that “it would be
nice if they picked up at the house and that would be, you know, especially if there's a lot of kids in the
same neighborhood, you know, it just would be closer.” Another suburban parent located in the same
town said that after enrolling their son in the school choice program it had become inconvenient
because they now have to like bring him toa bus stop instead of just sending him, you know, to the
corner.” When asked if their son could walk to his new bus stop for the magnet school, they noted that
even though it was geographically close it wasn’t walkable due to a forest in between their home and
the stop making the distance much farther than the distance as the crow flies.
Some parents were specifically concerned with the safety of stops location or the walk to them. One
Hartford parent described their child’s stop as particularly unsafe in “a really bad area. Drugs. It's
prostitutes... It's just really kind of hidden. And kind of dangerous." The same parent said she didn’t “feel
like the bus system really considers location when they're putting bus stops” and that her daughter had
been followed home one night from the stop, prompting the parent to demand a stop change.
Worries about safety, particularly early in the morning or on later afternoon stops, were not unique to
Hartford parents. A parent of a suburban student attending a magnet school noted that even though
they had a stop “right down the street...I still don't have a walk there because it's dark at night and on
the corner. Similarly, another parent specifically worried about her children’s stop being across a busy
road in the dark. In her words the walking trip was “unsafe because of the time that they [her children]
have to be there.” She also informed us that the public library that had served as a safe spot for her
children had been closed.
Parents worried about stop safety were sometimes concerned about busy roads at or on the way to the
stop. One parent with multiple children enrolled in magnet schools said she worried about one of her
kids crossing a particular busy road if they wanted to walk to their stop. Another parent whose child’s
bus stop was in a different location along the same major road noted that there were “no sidewalks on
that street,” and that in order to wait for the bus, they had to rely on pulling into a driveway off the road
because there was nowhere else to stand or wait.
In some cases, parents were specifically inconvenienced by the location of a bus stop in terms of its
impact on their routine and work schedule. One suburban parent said it was “so hard to get them there
9
The families we interviewed did not always distinguish between centralized, neighborhood, or home bus stop
locations. Thus, we use the generic term “bus term” throughout.
14
[the bus stop] and then back in the morning or in the evening trying to pick them up is difficult.” Another
parent of an elementary school student said she’d had problems with the centralized stop and her own
schedule, saying “when parents are waiting at a bus stop, that is not their own home or not in their
neighborhood, that they might have, like scheduling a conflict if the bus is even 10-15 minutes late.”
Some parents, even though most remained in choice, said they preferred the transportation options
offered by the traditional public school system, or that these options were more convenient to them.
This appeared to be particularly the case when local schools offered either neighborhood or door to
door transportation, and the magnet program limited them to stops that were further away. One parent
of multiple Magnet Choice students said that their local district would “pick the kids up in a reasonable
location where they could walk to,” as opposed to the magnet transportation stop, which was too far
away to travel by foot. Similarly, another parent in the same district said that when her child attended
the local school system, “the bus would pick her up right at the corner of where we live, so it was much
easier for me to be able to maneuver getting her to school, picking her up.”
Some parents much preferred their local transportation options but were still intent on keeping their
children enrolled in a magnet program school. However, other parents preferred their local options for
transit so much that they had withdrawn or planned to withdraw their students from a magnet after
graduating the school they had been in. One suburban parent who had chosen not to pursue further
magnet enrollment after her children finished an early childhood magnet said that this decision was
“partly also just because they're in a town school, we get the luxury of, like, a very convenient pickup
spot.”
In fact, some parents expressed regret that they had enrolled their children in school choice in the first
place due to these issues. One suburban parent said that due to the difficulty of the distance to the bus
stop, along with other issues with the transit system, that if she was redoing her school choice she
“definitely would have kept my kids in [Local District], like in the public school, because it would have
brought a lot less stress to my family.” Another parent in a similar situation said that looking back at her
school choice decision she “would definitely choose to have just kept them in the school where they
could use the public [school] transportation for our family. It would just make it so much easier.”
One Hartford parent with a student attending a magnet school worked to get a bus stop changed saying
she "fought for this one, but it's at the corner now of our streets. It took me a long time to get that bus
stop.” In a few different instances the bus stop was convenient for parents when it was near or in a
childcare center that coordinated with the choice school. This parent mentioned, “I bring them to a
daycare in town. And then from the daycare, they're bused to school.”
Parent Adaptations and Personal Transportation
School transportation is significantly less accessible for Hartford Open Choice parents when a parent
does not have personal transportation. A large majority of parents we interviewed (68 percent) drove
their children either directly to school or to the bus stop each day. Whether driving was a necessity or
preference among parents, the task often brought on burdens. For example, one Hartford parent who
sends her children to Open Choice and magnet schools said, “So the past three years, they rode the bus.
Well, yeah, except for last year I had so many complications with one of them that I started driving them
15
to school most mornings. And then this year I am driving them to school.” She explained that their
family made the decision to move to Hartford to, “have more education choices and so that [their]
transportation wouldn't be that long.”
One suburban parent spoke about her children having school disruptions from car issues, noting that
they “went through a period last year where our car broke down and like I said, it's not walking distance,
so CREC didn't really offer any other, you know, alternative in terms of getting them to school. So they
missed maybe like a week of school. So it's kind of inconvenient.” She said that when she reached out,
she was given few options. “When I didn't have a car, there was just really no options. And the school
had no resources for me, either. They told me to maybe take a Uber or do this or do that. But they didn't
like help with that in any way. She said that “even though they say every kid should have the choice of
what school they want to attend, it's not really a choice if you don't have a car,” which echoed the
sentiments of multiple parents.
Another parent expressed similar worries about potential issues with a family vehicle, saying that in the
case of a breakdown, “I would try to either have someone else come and help me bring him, or maybe
call an Uber. Those are the only options I can think of. But if it was more of a long-term thing, I would
probably have to put him back into the [District] system and take him out of the CREC program.Again,
this echoed parent sentiment that without car transit, many students would be unable to participate in
school choice. A third parent in a similar situation was forced to circumvent the bus system entirely and
take her child to school each day. She expressed that she
...felt like that's the way to shut out some of the parents who honestly have no, no option. There
is no option for me. Other than to take her myself. Because there's absolutely no way, based on
my schedule, (one), (two) based on where I live, for us to be able to get her bus to school, it's
just no option. And then on top of that. I'm being told, well, you'll either have to take her or she
can't be in school.
A theme that appeared multiple times was that parents had to leverage connections and make other
personal adaptations in order to navigate gaps they perceived or experienced in the transit system. One
parent said that in inclement weather they “have a relationship with one of the businesses in [Local
Mall], and so we've always told him that, you know, at the very least he could go into that business to be
warm or if it's snowing and you need to wait for us. A Hartford parent explained, “it was like a couple of
times where their bus would, like, completely miss their stops or I would not even get a phone call. The
only way I would even know sometimes would be from my aunt who actually worked at the bus
company.” Through these parents' adaptations, they were able to alleviate the gaps they experienced.
After School Extracurriculars
A handful of parents mentioned that they were disappointed that extracurricular activities at school
were inaccessible to their children due to lack of transportation. A parent whose child was enrolled in
Open Choice raised the availability of after school activity as a potential issue of equity, saying that “if
you ask any of the Hartford family parents about our kids, especially the kids, they would tell you that
there are barriers to their participation [in] after school activities. So not equitable.” She specifically
spoke about the possibility of Hartford students missing out on opportunities for developing
16
connections with peers in the district through after school extracurricular activities, such as clubs or
sports. She also worried that the system as it was “singled out” Hartford students in Open Choice noting
that when they missed out on these opportunities “then everybody also knows who the Hartford kids
are.”
Other families who were offered transportation for after-school activities expressed their frustration,
mentioning that the bus was often unreliable. One suburban parent explained,
So there was an after school bus for certain after school activities. Horrible. Several times the
bus didn't show up to pick them up from the after school activity. So I was waiting at the bus
stop and I was texting her and calling her and she's like, no, there's no bus here. And I'm like,
hey, I'm just going to come pick you up and then when I got to the school, the coordinator
would be like, I have no clue where the bus is. The bus never came, or the bus was supposed to
pick them up at 4. Sometimes it would show up at like 5. I just got in the habit of just picking her
up.
Another suburban parent said, “His bus sometimes does not pick him up from the school until as late as
5:30 or even 6:00. And so you can imagine he's the last stop... So we when he has after school activities,
we have no idea, really, what time he's actually going to arrive.”
Furthermore, some of these parents brought up their concern that transportation interfered with local
activities or extracurriculars because buses showed up late. One suburban parent said it was hard to get
her daughter to certain programs and that “she's missing out on stuff because you don't know when the
bus is going to be early or when it's going to be late. Another parent noted that “the only thing that is
really hard for us [about transportation] is after school sports.” They specifically singled out that their
town had only one late bus for all students in Hartford magnets, and so if an activity’s schedule did not
align with that one bus, they would have to miss it.
Communication
Approximately half of the parents (48 percent) mentioned that the communication system between the
transportation service and parents was unreliable. One Hartford parent said, “So like sometimes we get
the call, and it would say 10 to 20 minutes from your scheduled stop time, but it was already 10 to 20
minutes from our scheduled stop time.” Some parents sought alternative ways to receiving better
information about where their child was, for example, getting contact information of the bus driver. A
suburban parent said,
They finally found a permanent driver that I exchanged phone numbers with so that way we
could actually communicate, and she would say, ‘Hey. I'm running late today’ and then I would
get on the phone and I would call the bus company and say, ‘how come my driver can tell me at
7:00 that she's going to be late, but you are not sending out a notification. She was really good
about at least communicating and that's that was my biggest issue with the transportation.
There was no communication. We would be sitting there, and it would be 7:30 and we would
not get a notification.
17
Other parents felt the need to buy students devices such as phones or GPS watches to help track them.
A Hartford parent said that the bus was consistently outside of that window, enough that none of us
came at the beginning of that window. It was consistently outside of that window enough that I bought
a GPS watch for my kids, so I would know where they were. There were days when they would be 40
minutes late. Incidentally, vehicle GPS tracking services are provided by two of the three RSCO
Transportation bus companies, but no parents in our sample referred to it.
A Hartford parent mentioned that they tried to reach out to the bus company to talk about how to
improve communication. However, they were unhappy with the outcome.
[W]hen I called the bus company and I asked to sit down with somebody to see if we could find
or create a possibility, I felt really shut down when they were like, No, we don't have anybody
visit because of COVID.This was last year when I feel like it was getting to be mostly just the
choice of a company, not because of COVID. So I guess, yeah, I felt really shut down. I felt not
supported. I felt like there wasn't any way they were going to help find possibilities for my
children.
Sacrifices for School Opportunities
Despite 71 percent of parents communicating that they had some concern with school transportation,
many felt the education their children were receiving outweighed the challenges they were facing. One
suburban parent said,
I want to say yes to transportation like is a huge benefit, but... [I’m] making all these sacrifices to
get my kid to school so that they can have a better experience. It was hard those couple of
years. The thing is, is that my kids really need to go to that school. They do not fit in in our local
school and the magnet schools are truly a blessing to my family and so to get them there, I
would do anything but at the same time, transportation is such a huge benefit to my family. Like
truly... this is the best.
A Hartford parent who sends their children to an Open Choice school mentioned,
If there was no transportation, I would have to consider even moving to [Town]. Which I would
do because I love the school system. But maybe they would be going to a different school that's
a little bit closer. So yeah, if there was no transportation, they probably wouldn't be going to
that school.
Twenty-two (22) percent of parents expressed positive feelings about their bus stop location, mainly
parents in suburban districts with door-to-door or nearby bus stops. A suburban parent noted, “it
certainly made it easier because we live in the same town as the magnet school. We get door to door
transportation.” Likewise, another parent said, “It consists of us walking to the bus stop at the end of
our mailbox and the bus meets us there.”
However, for a small number of parents we spoke to (5 parents), their experience with transportation
was so negative that they decided to remove their child from the Open Choice or magnet school
program. A suburban parent mentioned,
18
We were in a position where we had to choose whose education is more important and it's a
horrible way to put it but, we have one son, our oldest, who's very computer technology design.
He flourished. He's in AP classes. I'm like this is important for him. He needs this type of
education. Our youngest. He's more about community and sports and he'll end up going to a
trade school. So I'm like, alright, you know having him back in [Local District] won't kill him for
the next three years.
Another suburban parent mentioned that she had made the choice to not continue enrolling her child in
magnet schools in part because they're in a town school, we get the luxury of... a very convenient
pickup spot.” She specified that this was a major factor in her decision process. Another parent living in
the same town said that she specifically pulled her kids from the magnet school program “because of
the transportation issues.”
Conclusion
Interviews with parents revealed students and family experiences with school transportation. We found
that the themes most frequently raised by parents included bus stops, parent personal adaptations and
transportation, after school extracurriculars, communication, sacrifices for school opportunities. Key
issues that emerged from the findings include, stop distance from home, safety of bus stops, need for
parent adaptations to alleviate transportation access issues, unreliable busing for extracurricular
activities or no busing for extracurricular activities, and an unreliable transportation communication
system. We provided recommendations at the end of the report to address these problems.
Limitations of Qualitative Study
The findings from parent interviews are based on parents who were willing or able to speak
with us. We made extensive efforts to recruit parents to the study, but only a small fraction of
parents we randomly contacted ended up participating. Thus, we cannot be certain the parents
in our sample are necessarily representative of all parents whose child attends a magnet or Open Choice
school. Finally, parents do their best to explain their experience with transportation to and from school,
but this is based on their view and absent from the perspective of the school. Nonetheless, the parent
views represented here revealed some distinct patterns overall.
19
Study 2: Bus Complaint Logs
Bus Complaint Logs
The Capitol Region Education Council (CREC) serves as the transportation contractor on behalf of RSCO.
During the year, the CREC Transportation Office fields travel-related complaints from parents. Parents
can submit their complaints by phone, email, or an online web form. Customer service representatives
record complaints in a database, assigning each a ticket ID for record keeping and ongoing resolution.
Each ticket also includes the date of the initial complaint, the school in which the student attends, a
brief description of the complaint and any actions in response, and a code categorizing the nature of the
complaint. The latter represent predetermined categories such as “bus/vehicle driver,” “central stops,”
and “excessive lateness;” the categories are either selected by families using a dropdown list on the
online form or, in the cases of phone calls or voice messages, are assigned by CREC.
Analytic Approach
RSCO provided us access to the 2022-23 complaint dataset, which contained 1,285 unique complaints.
For each complaint there were typically 2-3 records documenting correspondence with parents or other
transportation personnel. For instance, customer service representatives would issue notes such as
“returned call to family” or describe in detail their actions in response to the complaint. Often the
records indicated the situation was resolved with no further action required. In total there were 6,148
records for the 1,285 complaints. Nearly 85 percent of the 1,285 complaints were resolved or closed
with three or fewer ticket documentations.
We descriptively analyzed complaints by ticket, generating frequency distributions for key fields such as
complaint ticket source, ticket date, and complaint type.
Findings
More than three-quarters (77.4 percent) of the complaints were submitted by phone (Table 2), followed
by the online web form (15.5 percent), and email (4.0 percent). Complaints were also disaggregated by
the students’ school. Magnet schools that had at least 20 complaints are listed in Table 3. The list is
ordered by “complaint per student ratio” to account for school size (but not the number of students
bused to that school). Some schools received relatively more complaints than others. Reggio Magnet
School of the Arts had the highest complaint per student ratio,
10
followed by Ana Grace, Museum
Academy, Aerospace and Engineering Elementary, and Glastonbury-East Hartford School for Global
Citizenship. These schools serve elementary grade levels, which may partly explain their higher rate of
complaints; that is, parents and guardians of young children may be more apt to be concerned for their
child’s welfare.
10
Incidentally, we heard from Reggio parents in our interview sample that Reggio transportation worked
well for suburban families living nearby, and not as well for those living far away from the school.
20
Table 2. Complaint Submission Source, CREC Transportation Database, 2022-23
Source
N
%
Call, Inbound
995
77.4
Call, Outbound
28
2.2
Email, Inbound
51
4.0
Email, Outbound
3
.2
In Person
3
.2
Web Form
199
15.5
Letter, Inbound
1
.1
Other
2
.2
Skype
1
.1
Voice Mail, Personal
2
.2
Total
1,285
100.0
Table 3. Transportation Complaints by Student Magnet School Destination (minimum 20 complaints),
2022-23
Magnet School
No. of
Complaints
School
Enrollment
Complaint
per Student
Ratio
Reggio Magnet School of the Arts
70
514
0.14
Ana Grace Academy of Arts Elem Magnet
96
863
0.11
Museum Academy
57
513
0.11
Academy of Aerospace and Engineering Elem
64
578
0.11
Glastonbury-East Hartford Elem Magnet
50
495
0.10
Internat’l. Magnet School for Global Citizenship
30
494
0.06
University of Hartford Magnet
30
510
0.06
Environmental Sciences Magnet at Mary Hooker
29
556
0.05
Discovery Academy
27
520
0.05
Academy of Aerospace and Engineering
39
768
0.05
Kinsella Magnet School of the Performing Arts
22
465
0.05
Classical Magnet School
20
445
0.04
PSA Civic Leadership High School
21
496
0.04
Academy of Science and Innovation
30
756
0.04
Sports And Medical Sciences Academy
20
530
0.04
Two Rivers Magnet Middle School
24
642
0.04
MLC for Global and International Studies
26
709
0.04
CT River Academy
21
664
0.03
Hartford Magnet Trinity College Academy
27
963
0.03
21
Table 4 shows the categories of complaints that were either selected on the online form by the family
member or, in cases in which a phone call was received, were issued by CREC Transportation customer
service representatives. Most of the complaints fell under “bus/vehicle driver,” “late or no notification
of delays,” and a category labeled as “other.”
In reviewing the descriptions of the complaint entered for each ticket, we noticed some did not
necessarily align with the reason code. Part of the issue could be that some of the categories overlapped
with one another (e.g., “central stops” and “stop location”). Another could be that certain complaints
implicated more than one reason code but only one could be selected. "Stop location” and “central
stops” were not prominent concerns, at least as recorded in the complaint database. Incidentally, this
contrasts with what we learned from the parent interviews, where bus stops were raised as a significant
issue. We suspect the complaint system categories may be diffused to the point where stop location
could be classified
11
In other words, bus stop complaints may not all be captured under the “stop
location” and “central stop” codes.
11
Here is an example from a complaint ticket classified as a ”bus/vehicle driver” problem: “Parent called stating
that the driver is dropping her student off in an unsafe location at the intersections of Lyme St & ….”
22
Table 4. Complaint Codes Recorded in CREC Transportation Complaint Database, 2022-23
Complaint Code
N
%
Bus Monitor
53
4.1
Bus Schedule
95
7.4
Bus/Vehicle Driver
502
39.1
Central Stops
13
1.0
CREC Athletics
1
.1
Excessive Lateness - AM or PM Transport
27
2.1
Late or No Notification of Delays
255
19.8
No notification received of route or stop change
16
1.2
No Response
6
.5
Other
142
11.1
Route Too Long
20
1.6
Stop Location
62
4.8
Student
93
7.2
Total
1,285
100.0
Unsurprisingly, most complaints were issued early in the school year (Table 5 and Figure 1), a time when
buses and families are adjusting to the new routine. As the year progressed, complaints tapered off.
Figure 2 shows complaints by month and by complaint type; bus/vehicle driver” complaints, although
declining over the school year, remained an issue throughout to a degree.
Table 5. Complaints by Month of Occurrence, 2022-23
Month
N
%
August
29
2.3
September
356
27.7
October
235
18.3
November
122
9.5
December
90
7.0
January
110
8.6
February
89
6.9
March
100
7.8
April
59
4.6
May
62
4.8
June
21
1.6
July
12
.9
Total
1,285
100.0
23
Figure 1. Complaints by Month of Occurrence, 2022-23
0
5
10
15
20
25
30
Percent
Month Complaint Filed
24
Figure 2. Complaint Type by Month of Occurrence, 2022-2023
More on Complaint Descriptions
We did not analyze each recorded complaint as there were a substantial number. Below we offer two
examples of what the complaints look like in the log. We use them as illustrations to provide a window
into the system. Coincidentally, they also capture instances of some of the more common concerns we
learned from the parent interviews.
The first is a complaint a parent submitted through the online form and, incidentally, selected the “bus
schedule” category.
12
The parent has two children attending different magnet schools and parent
expresses a concern with one of the stop locations. According to the record logs, the parent submitted
an appeal to the RSCO Transportation email and offers a possible solution.
The pick up/drop off location is 3.1 miles and 45-1hr walk through extremely busy roads and intersections
i.e [lists 2 intersections]. I do not have a car to drive him. My son leaves at 5:15am while it is still dark
outside. Sidewalks are not 100% present, there are large hills, the weather is bad and this is extremely
unsafe for any child to walk through. This leaves my child at risk of car accidents hypothermia, asthma
attacks and kidnappings. Please allow us to work together to come up with a location that fits the current
route while also ensuring that my child is safe to get to his bus stop and home each day.
12
Later, a customer service representative changed this to “stop location” and issued a new ticket ID.
25
The parent subsequently appeals the stop location and requests a new stop.
I've submitted numerous requests for route changes as I do not have a vehicle to transport my children to
a bus stop 2-3 miles away.
...
Requesting stop location [streets A & B]
The second example is a complaint called in by a parent concerned with a late bus and the timeliness of
receiving a notification. Because it was called in by a parent, the description was recorded by a customer
service representative.
Parent [Name] called in looking for an ETA to student [Name] bus stop location, parent stated the bus is
constantly late without receiving notification in a timely manner, sometimes she does and sometimes she
doesn't. at 8:09am the bus company put in...
East, 8:09 AM
[XXX XXX] BUS [XXX] Spare driver will be at 1st stop in 4mins. Running 5-10 mins otw
It is about a 15min gap between stop locations Parent pick up time is for 7:52am, parent opted to bring
her students to school instead of being late parent wanted to put this complaint in hoping for some
resolution regarding this matter.
The customer representative documented an additional response:
I reached out to the parent of [student] at [phone number] to find out how's things have been going with
her student's route, parent didn't answer so I left a detailed voicemail explaining to reach out to the
company with any other updates or concerns regarding her student bus route.
Summary
The complaint log system offers parents a way to immediately share transportation concerns with the
bus company. In our review of a random sample of about 10 percent of the complaints, it appears CREC
customer service representatives often can react in a timely manner and are able to resolve the issue. In
some instances, communication was delayed or impeded, likely due to having to navigate through
several layers of the bus ecosystem. The log data can provide useful information to transportation
personnel looking to improve upon the transportation experience for students and families. But the
feedback from the complaint logs is only as good as the quality of records. There seem to be areas for
improvement in recording the nature of the complaints in the system.
26
Study 3: Student Travel Times to Schools
and Bus Stops
Study 3A: Student Travel Times to Schools
We conducted a series of geospatial and statistical analyses using bus schedules, bus routes, and
student address data provided by the RSCO Transportation Office.
Analytic Approach
Using bus schedules for students enrolled in a Choice program, we generated expected bus ride times.
Although the expected times do not represent actual times of travel, we assume these are reasonably
close estimates of actual time traveled to school. The initial dataset included 12,182 students --10,359
enrolled in magnets and 1,823 in Open Choice in the 2022-23 school year. For our analysis, we examined
students attending interdistrict magnet schools operated either by Hartford or CREC. The 1,134
students enrolled in magnets overseen by other operators were excluded from the analysis.
We conducted descriptive analyses of the expected bus ride times by student resident group and choice
program, as well as sending district for magnet students. Because bus stops were a major concern
among parents we interviewed, we also used the dataset to estimate student distance to their stops to
evaluate theoretical walkability for students. Estimates of walking travel time to stop was produced for a
representative sample of students and used to assess the feasibility of walking to their bus stops from
their homes.
Findings
Expected Bus Ride Times
Table 6 displays expected bus ride times across a variety of groups. The expected median bus ride from
stop-to-school for magnet students was 38 minutes, with suburban students exhibiting a slightly longer
median ride time (39 minutes) relative to Hartford students (36 minutes). Hartford Open Choice
students had the longest median expected ride time at 44 minutes.
13
We also disaggregated data by distinct ride time thresholds. That is, we deemed ride times over 30
minutes a ”long” ride and any ride over 60 minutes to be ”very long.” Among Open Choice students,
88.1 percent had a long expected bus ride, and 18.6 percent had very long expected ride time,
13
Cordes et al. (2022) noted that there is no universal agreement on what constitutes a long ride for students
traveling to school, and so created their own definitions of long (45-60 minutes) and very long (>60 minutes).
27
compared to 72.3 percent and 11.0 percent respectively of students in magnet schools. For students in
magnet schools, suburban students generally take longer bus rides with a median of 39 minutes
compared to 36 minutes for Hartford students.
Table 6. Expected Bus Ride Times by Choice Program and Student Resident Group, 2022-23
Choice Program by Resident
Group
Median
Expected Bus
Ride Time
(Minutes)
Median
Absolute
Deviation
Percent of
Students
with Ride
Time ≥ 30
min
Percent of
Students
with Ride
Time ≥ 60
min
N
(known
bused
students)
Open Choice Hartford residents
44
10
88.1
18.7
1,823
Magnets
38
15
72.3
11.0
10,362
Hartford residents
36
8
73.3
5.2
3,726
Suburban residents
39
12
71.8
14.3
6,634
Hartford Host Magnets
37
10
68.8
7.5
3,649
Hartford residents
32
7
60.8
0.0
332
Suburban residents
37
11
69.6
8.3
3,317
CREC Magnets
39
11
74.2
12.9
6,711
Hartford residents
36
8
74.5
5.7
3,394
Suburban residents
41
13
74.0
20.3
3,317
These trends align with the location of most magnet schools in the metro Hartford area. Suburban
students entering the Hartford metro to attend magnet schools and Hartford students leaving Hartford
to attend Open Choice schools in the suburbs would unsurprisingly have longer travel times than
students traveling within metro Hartford. In addition to having the lowest median travel time (32
minutes), Hartford residents attending Hartford hosted magnets show the narrowest distribution of
expected ride times, particularly when compared to Open Choice students and suburban residents
attending magnet schools. Interestingly, suburban residents attending Hartford Magnets had a median
expected ride time (37 minutes) below the overall median for magnet students (38 minutes) and below
the median for suburban students in CREC magnets (41 minutes) with a generally narrower distribution.
This might suggest that regionality plays a role in terms of which Hartford hosted magnets suburban
students apply to and attend, with some preference to schools located within the parts of metro
Hartford easily accessible from a student’s home district. On the other hand, CREC magnet schools may
have more students “passing through” the Hartford metro.
Figure 3 displays the distributions of estimated bus ride time by choice program and resident group
using box and whisker plots. While median scores are useful (column 2, Table 6), box plots allow for
direct comparisons across groups in terms of spread and central location of scores. The boxes represent
the middle 50 percent of the score distribution (i.e., 25th to 75th percentile). A horizontal line is typically
included within the box to represent the median, but our program did not produce it. We drew in lines
for Open Choice and Magnet distributions as an example. The distributions in Figure 3 are all positively
skewed, meaning most scores are bunched to the lower ride times side, and the remaining scores are
28
more dispersed all the way to the maximum ride time. The first two box plots show that Open Choice
and magnet students have similar spread and concentration of estimated bus ride times, although Open
Choice students have a generally longer ride.
Figure 3. Expected Bus Ride Times by Choice Program and Resident Group, 2022-23
In addition, expected bus ride time for magnet school students was examined by sending district. Table
7 shows all districts sending at least 30 students to magnet schools as organized by median expected
bus ride time. The variation across districts was considerable, and likely heavily explained by district
proximity to central Hartford, where the bulk of magnets are located. Students from Tolland, for
example, had a median expected ride time of 62 minutes, with 100 percent having a “long” expected
bus ride and 60.6 percent with a very long” bus ride time. On the other end of the spectrum, students
from West Hartford had a median expected ride time of only 24 minutes, with 39.4 percent having long
bus rides and 7.4 percent having very long ride times. Notably, while West Hartford had the lowest
median expected ride time, Plainville (median 35.5 minutes) and Wethersfield (median 26 minutes) tied
for the lowest percentage of students with very long expected bus rides with 0 percent. Students from a
given district likely would be attending different magnet schools, and as such some variance at the
school level should be expected within districts.
29
Table 7. Expected Bus Ride Times for Suburban Magnet Students by Sending District (Sending 30 or
more students), 2022-23
Sending District (30 or More
Students)
Median Expected
Bus Ride Time
(Minutes)
Percent of
Students with
Ride Time ≥ 30
min
Percent of
Students with
Ride Time ≥ 60
min
N
Berlin
53
77.1
43.8
48
Bloomfield
33
61.7
9.8
379
Bristol
55
97.0
36.5
233
Cromwell
37
94.3
17.0
53
East Hartford
35
60.8
8.8
1,240
East Windsor
38
67.7
6.5
31
Ellington
46
91.1
26.8
56
Enfield
49
80.4
30.4
204
Farmington
32
58.8
2.9
34
Glastonbury
47.5
76.9
15.4
52
Hartford
36
73.3
5.2
3,728
Manchester
37
72.4
10.9
997
Middletown
49
97.2
20.6
141
New Britain
43
83.6
11.6
1,116
Newington
32
56.9
12.7
102
Plainville
35.5
78.3
0.0
46
Rocky Hill
37
58.1
11.3
62
South Windsor
41
66.7
13.3
105
Southington
39
95.2
14.3
63
Tolland
62
100.0
60.6
33
Torrington
55
100.0
35.9
78
Vernon Rockville
42
74.4
17.4
207
West Hartford
24
39.4
7.4
203
Wethersfield
26
36.4
0.0
129
Windsor
29
47.0
3.8
419
Windsor Locks
42
75.3
12.9
93
Figure 4 shows box plots by sending district. Differences in both spread and central tendency are evident
across districts. As noted above, the location of sending districts in relation to most magnets likely
explains some of the ride time differences.
30
Figure 4. Expected Bus Ride Times for Suburban Magnet Students by Sending District (Sending 30 or
more students), 2022-23
Figure 5. Percent of Magnet Students with Long and Very Long Bus Ride Times in Suburban Districts
(Sending 30 or more students), 2022-23.
How do Estimated Bus Ride Times Compare to Traditional Public Schools?
To get a sense of how ride times for RSCO students compared to students attending their local (non-
choice) schools, we generated expected bus ride times for students in a small sample of districts. We
selected New Britain, South Windsor, and Vernon/Rockville to represent a range of geographies in the
Sheff region and because their district bus route data were publicly available. Using 2021-22 bus route
information for the afterschool drop-off, we calculated the time between students’ scheduled departure
from school and scheduled drop-off time at their bus stop. A possible limitation to our measure was that
31
we mathematically assumed one student per stop location, which may be an inaccurate measure.
Overall, however, this should have minimal impact for the purposes of comparing ride times on a basic
level. New Britain students traveling to a district school had a median expected travel time of 12
minutes, compared to 43 minutes for New Britain students traveling to magnet schools. South Windsor
students transported within district had an expected median ride time of 14 minutes, compared to 41
minutes for students going to magnets. Finally, Vernon/Rockville district students had a median ride
time of 17 minutes compared to 42 minutes for magnet students. In all three cases, attendance in
magnet school programs represented a substantial increase in expected bus ride times for students; this
is thoroughly consistent with the research literature, which demonstrates choice students experience
longer bus rides than what they would have attending their neighborhood-assigned district school
(Corcoran, 2018).
32
Study 3B: Student Distance from Home to Bus Stop
Based on “bus stops” as a prominent theme that emerged from parent interviews, we analyzed the
theoretical distance students would have to travel to reach their bus stops from their homes.
Analytic Approach
We selected a representative sample of 2,500 students from the total population of students attending
magnet programs and Open Choice in the 2022-23 school year to analyze the distance between student
home locations and their bus stops. The magnet sample was created starting by randomly selecting 500
suburban students in Hartford Host magnets using a random number generator. The resident town for
each student was noted, which then allowed matching to a random student within suburban students in
CREC magnets from the same town, creating a sample with 500 students with the same distribution of
sending districts between CREC and Hartford magnet students from outside Hartford. This was done to
decrease the possible effects of individual town on the measurement of stop distance. From there, a
sub-sample of 1,000 students from Hartford attending magnet schools was created using 300 students
from Hartford magnets and 700 Hartford students from CREC magnets. The imbalance in this sub-
sample was due to the limited number of students receiving school bus transportation from Hartford
attending Hartford Host magnet schools (a total of 332 students). Finally, 500 Hartford-resident students
from the Open Choice program were randomly selected.
A code process using an embedded version of the Google Map Application Programming Interface (API)
was used to determine the walking travel time between a student’s home and their bus stop
14
. The
macro calculated the shortest-travel time walking path between the two locations using the API given
average conditions, not factoring in time of day or conditions such as weather, traffic, or other factors.
Due to the computer power required, estimates of walking times were completed in limited quantities
over time to avoid “overheating” the system. In some cases, no viable walking path could be mapped
most likely due to random computational error. In total, just over 1 percent of cases produced this error.
These cases were dropped from the sample and their spot was resampled randomly from previously
unselected cases in the same sending town to preserve the representative sample.
Findings
The results of the analysis show a wide range of travel times for students, with the widest difference
being between Hartford students (in both magnet schools and Open Choice) and suburban magnet
students (Table 8). This is to some degree unsurprising as the nature of Hartford as a city makes
walkability to stops more likely. Hartford students in Open Choice had the lowest median walking
distance to their stop with 2 minutes, with a relatively tight distribution of travel times as can be seen in
Figure 8. Hartford students in magnet schools had a median walking travel of 4 minutes, with a slightly
higher range of distances.
14
All identifying student data outside of home and stop location were removed from this data before processing.
Address data remained solely within the data sheet and was internally processed by the embedded Google Maps
API.
33
Table 8. Median Estimated Walking Time for Home-to-Bus Stop Travel by Choice Program and Resident
Group, 2022-23
Program and Resident Group
Median Expected Walking Time
Between Home and Bus Stop (min)
Median Absolute
Deviation
Open Choice Hartford residents
2
1
Magnets
7
5
Hartford residents
4
2
Suburban residents
35
24
Hartford Host Magnets
7
4
Hartford residents
6
2
Suburban residents
78
65
CREC Magnets
7
5
Hartford residents
4
1
Suburban residents
27.5
16.5
Suburban students in magnet schools had a median walking travel time from home to their bus stop of
35 minutes. The range for these students varied a sizeable amount. The distribution for these times had
a heavy positive skew with several extreme outliers on the high minute end. The median absolute
deviation of 24 for suburban residents in magnets reflects this. The mode for magnet school students
was 15 minutes of walking travel while the mean was 59 minutes. A 15-minute walk might be
considered reasonable depending on the age and circumstance of some students, however 54.5 percent
of cases of suburban students in magnet schools had a walking travel time of 30 minutes or more,
indicating that for the majority of these students walking to their bus stop would be at the very least
difficult if not unfeasible entirely. Again, the data does not indicate whether a given student typically
walks to their bus stop or not in reality, however we presume that excessive estimated walk times
would ostensibly require students to be driven.
Within the sample, there emerged extreme cases where the theoretical walking travel time between a
student’s home and bus stop was multiple hours, sometimes even four or more. These cases were
directly examined using the Google Map API macro to view the detailed walking directions. In many
cases, these extreme travel times were the result of some obstacle (e.g., highways, unpathed woods,
bodies of water) in the way of the direct path to the bus stop. It is perhaps more useful to consider these
high travel time stops as being just effectively unwalkable, rather than considering the actual duration of
the trip itself. When controlling for these cases, the median absolute deviations of the sample does not
shift substantially, suggesting that the high variability shown in for suburban residents in magnet schools
persists regardless of outliers.
The data show that many suburban students attending magnet schools live outside of a feasible walking
distance from their bus stops, which aligns with the complaints about stop distance and walkability that
emerged from interviews with parents. Based on these factors, it is most likely reasonable to say that
the median suburban family cannot participate in Magnet School Choice without consistent access to a
car or other form of personal transportation.
34
Figure 6. Estimated Walking Time for Home-to-Bus Stop Travel by Choice Program and Resident Group,
2022-23 (cases with estimated walking time > 250 minutes removed)
It is perhaps useful to frame student home-to-bus stop travel time alongside their expected bus stop-to-
school bus rides. Considering residents of Hartford are likely to be able to walk to their bus stops, adding
the medians of the expected ride times and estimated walking travel times gives us an approximation
for total transit time, not including waiting at stops. Hartford residents in open choice have a median
expected bus ride of 44 minutes and a median home-to-bus stop walking travel of 2 minutes, for a total
of 46 minutes expected for transit each way. Hartford students in magnets have a median expected ride
time of 36 minutes with an estimated stop distance of 4 minutes for an estimated 40 minutes of total
transit each way. The median suburban student attending a magnet program would have an estimated
walking distance to their stop of 35 minutes and a bus ride time of 38 minutes indicating a total travel
time of 1 hour and 13 minutes without considering wait time at their stop. Considering a 35-minute walk
is outside the range of feasibility for many students, this estimate is likely high as many suburban
students are likely to be driven to their stop locations. With that said, students are often directed to
arrive at their bus stop up to ten minutes before and be prepared to wait ten minutes after their pick-up
time, possibly adding 20 minutes of waiting. It may therefore be reasonable to suggest the median
suburban student attending magnet schools dedicates an hour or more of their day to travel, each way,
assuming they take the bus.
35
Study 4: School Travel Times and Parent
Lottery Decision Making
We were interested in learning the degree to which travel distance to school was related to parent
decisions to accept or decline a lottery placement offer. Our analysis was similar to the expected bus
ride times above but uses a different dataset. The lottery data contained placement offer
accept/decline fields, which we linked to (theoretical) estimated travel times from student home
address. We also considered the estimated commute times by Open Choice receiving district and
magnet school sending districts. We did so to explore possible tendencies or patterns in estimated
commute times occurring at the district level.
Analytic Approach
We used RSCO lottery data to estimate driving time between the home location and school of 10,186
students who had received first-round offers to either a magnet (n=9,421)
15
or Open Choice school (765)
in 2022-23.
16
We linked these data to parent lottery offer decisions. We were also able to aggregate
lottery offer acceptance rates by the offered magnet school or Open Choice district for subsequent
analyses.
As in the prior analysis of student home-to-bus stop, we used a macro embedded form of the Google
Maps Web API. The Google Map API was used to determine the driving time without consideration of
time of travel or traffic conditions. The estimated driving time from student home to placement school
served as a rough proxy for commute time. Although it does not represent the actual commute time a
student would experience traveling to school, it provides a hypothetical proximity-to-school measure
parents may consider as they make their decision on the placement offer. Our measure is superior to a
simple distance-to-school measure, such as a “as the crow flies” measurement or even distance in miles
by vehicle. Our measure takes into account typical driving speeds based on roads, stop lights and stop
signs, and the like. Another advantage of using estimated driving time is that it serves as a common
metric to compare parents; the drawback is beyond obvious walking distances/times, we do not know
how the child would eventually get to school by bus or car. Although most would likely be bused,
these are parent decisions based on their individual circumstances.
One would expect that bus stop locations would also play a role in parent decision making, although, it is
unlikely parents would know the bus stop location at the time of the placement offer. They also would
not likely know the bus route and its actual transit time. Another limitation of using estimated driving
time is it does not accurately estimate bus ride times. Bus ride times are also a function of the number
of stops, student travelers, and traffic congestion at the times of travel. For longer estimated travel
times by the method we used, the longer the hypothetical bus ride due to presumably more stops along
the way. Buses also simply travel more slowly than cars. Thus, our estimated driving time measures are
used as a means to an end that is, to assess the relationship between proximity to school and seat
15
One application was missing necessary information, which led to our final n=9,421.
16
We could not estimate driving time for 15.3 percent of students.
36
offer acceptance rates). Nonetheless, they have some value in isolation, so we present our commute
estimates in Appendix B, along with comparisons between Open Choice receiving districts (Figure B1),
magnet school sending districts (Figure B2), and magnet school offered (Table B1).
We conducted a separate set of analyses on the 2022-23 lottery data sample to examine relationships
between observable factors that could influence parent decisions. We began by looking at bivariate
relationships two variables at a time. For instance, we examined the relationship between school rank
and parent decisions. Subsequently, we employed a binary logistic regression to assess the relationship
between parent decisions and a host of predictor variables simultaneously. In the model, the dependent
variable was the decision by parents to either accept or decline a magnet seat (declines also include
administrative declines). The variable was coded as 0=decline and 1=accept. The model is designed to
estimate the independent relationship (i.e., “effect”) each predictor variable has on the outcome, while
controlling for the remaining predictors. Our model included five predictor or explanatory variables:
school preference ranking on lottery application, SES tier (higher score means higher income), reduced-
isolation student status,
17
Hartford resident status, grade span of the choice school, and estimated
driving time.
Findings
There was a moderate in strength, inverse relationship (r= -.453) between estimated median driving
time and magnet seat acceptance rate per sending town (Table 8) as visualized in Figure 7. This suggests
that, generally, when travel time increases, acceptance rates decrease but not in a perfect linear
fashion. Students who accepted their placement had a median estimated travel time that aligned with
the general sample at 15 minutes; however, parents who actively rejected their placement offers had a
median estimated driving time of 17 minutes. When including cases of “administrative declines” with
decline placements, the median was again aligned with the overall sample median of 15 minutes. When
examining only cases with an estimated ride time of less than 30 minutes, the overall median was 14
minutes. Students who accepted placement had a median of 14 minutes under these conditions, while
students who declined their placement had a median of 15 minutes. For Hartford-resident Open Choice
students, there was a slightly smaller difference between students who accepted their offer (median =
22 minutes) and those that rejected (median = 23 minutes). There was no difference in median
estimated travel distance for either group when applying the cutoff of 30 minutes estimated travel time.
While suburban and Hartford magnet students who accepted placement offers had an overall lower
median estimated travel time than students who actively declined
18
(15 min vs. 17 min, respectively) this
difference was relatively small, especially in comparison to the between-school differences (Table 9) and
between sending districts (Table 10) for magnet students. Beyond this, when including cases of
17
“Based on the Hartford-resident demographics and the goal of reducing isolation, a “reduced isolation student”
is a student who identifies as White, Asian, American Indian, Alaska Native, Native Hawaiian and/or Other Pacific
Islander, or two or more of such races, and does not identify as Black/African American or Hispanic/Latino.” (p. 5 of
the Comprehensive Choice Plan)
18
We refer to ”active declines” as when parents indicate they are declining the offer. When a parent does not
respond to the offer, for whatever reason, RSCO considers that an ”administrative decline.”
37
“administrative” declines in with student declining offers directly, there was no difference between
medians of students accepting and declining placement. In addition, there was a substantially weaker
inverse relationship (virtually no relationship) between accepted offer rate and median estimated travel
time when grouped by school (r = -.055) compared to when grouped by town (r = -.453) (see Figures 7
and 8 below for a comparison of their visual relationships). There was also a slightly positive relationship
between the number of placements offered to students per sending district and the acceptance rate of
offers per town. Together, this may suggest that distance has a larger impact on the choice to accept or
reject offered placement in some districts compared to others.
The weaker relationship when sorted by schools could also be explained by the idea that certain schools
have strong regional draw. This could be viewed similarly to how in college athletics, top tier programs
often are perceived as having control over their local pool of talent when it comes to recruiting
prospective student athletes. In the case of magnet schools, it is possible that some schools have similar
pull for applicants in a given subsection of the capital region which would make travel time less of a
factor. For example, the Montessori Magnet at Annie Fisher School is located in the northwest of
Hartford near the West Hartford border. It may exert effectively regional draw for students in suburbs
north and west of Hartford, or who are along the route 291 corridor, for whom moderate differences in
perceived distance to the school might have little effect. On the other hand, the Montessori Magnet at
Batchelder might have higher regional draw for students towards the south of Hartford, such as
Newington, New Britain, or Rocky Hill. For students outside of these regional zones, distance may be a
larger factor in their decision making in terms of accepting or rejecting an offered placement, leading to
the relationship observed between sending district median estimated travel time and acceptance rate
despite the relatively small difference in median estimated travel times between students accepting or
rejecting placement.
In view of these factors, it may be reasonable to presume that consideration of travel distance occurs in
advance of parent decisions in the face of a placement offer. . It is certainly possible that distance is
considered by most families in advance of applying to schools, and that many families prioritize applying
to schools that are within the comfortable range of travel for their needs.
38
Table 9. Estimated Travel Time Between Student Home and Offered Magnet School Placement by
Offering School
Offering Magnet School
Median Estimated
Drive Time (min)
Offer
Acceptance
rate
N (Placement
offers)
Academy of Aerospace and Engineering
19
66.7
221
Academy of Aerospace and Engineering Elementary
17
52.3
320
Academy of Computer Science and Engineering
22
60.5
325
Academy of Computer Science and Engineering MS
11
72.5
531
Academy of International Studies
20
51.4
327
Academy of International Studies Elementary
14
69.5
302
Academy of Science and Innovation
19
64.3
258
Ana Grace Academy of the Arts
23
72.2
208
Betances Learning Lab
11
56.5
330
Betances STEM Magnet School
13
58.2
144
Breakthrough Magnet School, North
15
64.2
173
Breakthrough Magnet School, South
12
52.6
124
Capital Preparatory Magnet School
13
38.4
331
Classical Magnet School
12
39.9
165
Connecticut IB Academy
17
51.4
109
Connecticut River Academy at Goodwin
10
51.3
311
Discovery Academy
11
70.5
154
Early College Advanced Manufacturing
10
67.2
18
Environmental Sciences Magnet at Hooker
15
59.9
246
Glastonbury/East Hartford Magnet School
16
66.8
316
Global Experience Magnet School
24
72.4
178
Great Path Academy at MCC
15
77.2
165
Greater Hartford Academy of Arts (HD)
19
51.9
342
Greater Hartford Academy of the Arts
14
55.8
138
Greater Hartford Academy of the Arts High School
17
54.5
190
Hartford Magnet Trinity College Academy
13
57.7
250
Hartford PreKindergarten Magnet School
12
79.6
308
Kinsella Magnet School of Performing
13
64.7
299
Montessori Magnet at Batchelder
9
66.7
189
Montessori Magnet at Fisher
15
45.2
161
Montessori Magnet School (CREC)
16
53.3
177
Museum Academy
14
49.0
198
Pathways Academy of Technology and Design
11
70.7
157
Reggio Magnet School of the Arts
20
52.5
240
Riverside Magnet School at Goodwin
11
58.8
239
Sport and Medical Sciences Academy
14
60.3
207
STEM Magnet at Annie Fisher School
15
61.5
179
University High School of Science and Engineering
16
46.1
158
University of Hartford Magnet School
14
40.8
255
Webster Micro Society Magnet School
10
42.1
246
Wintonbury Early Childhood Magnet School
15
59.9
232
39
Table 10. Estimated Travel Time Between Student Home and Offered Magnet School Placement by
Student Location
40
Student Location (Locations with <10 Students
excluded)
Median Estimated
Drive Time (min)
Offer
Acceptance
rate
N (Placement
offers)
Avon
20
55.1
69
Berlin
24
35.0
40
Bloomfield
12
66.6
338
Bolton
20
57.1
21
Bristol
29
48.4
153
Burlington
33
62.5
24
Canton
31
36.4
11
Cheshire
32
45.5
11
Colchester
34
16.7
18
Collinsville*
33
18.2
11
Coventry
28
37.5
24
Cromwell
22
46.9
64
East Granby
17
64.7
34
East Hampton
30
35.5
31
East Hartford
11
63.6
767
East Windsor
22
40.0
30
Ellington
26
36.8
68
Enfield
23
39.4
127
Farmington
20
37.0
54
Glastonbury
13
43.2
162
Granby
26
50.0
24
Hartford
11
69.3
3748
Manchester
17
53.6
662
Marlborough
24
50.0
12
Meriden
27
34.7
49
Middletown
25
45.3
117
New Britain
19
55.6
606
New Hartford
38
63.6
11
Newington
15
43.2
125
Plainville
21
63.3
30
Plantsville
25
45.5
11
Portland
24
40.0
30
Rocky Hill
19
33.8
133
Simsbury
18
75.6
45
Somers
34
17.6
17
South Glastonbury*
17
66.7
33
South Windsor
15
42.1
242
Southington
27
36.7
49
Stafford Springs
35
30.0
10
Suffield
25
55.6
18
Tolland
27
47.4
19
Torrington
46
36.8
38
Unionville
26
36.4
33
Vernon Rockville
21
46.6
161
41
Wallingford
31
25.0
16
Waterbury
37
33.3
42
Weatogue
20
41.7
12
West Hartford
12
48.3
319
West Simsbury
20
66.7
15
West Suffield*
27
27.3
11
Wethersfield
11
41.1
151
Willington
37
58.3
12
Windsor
14
52.1
280
Windsor Locks
19
45.5
55
*Indicates a sending location with at least 10 students that is part of a larger town or district
42
Figure 7. Magnet Placement Offer Acceptance Rate by Estimated Median Drive Time per Sending
Location (Districts with fewer than 10 students offered placement excluded)
(Hartford noted by red square)
Figure 8. Magnet Placement Offer Acceptance Rate by Estimated Median Drive Time per Offering
School
43
Lottery Data Analysis (First-Round)
The lottery data set afforded us another opportunity to determine the extent to which travel to school
influenced parent decisions in the face of a placement offer to either a magnet or Open Choice school. A
multivariable model was developed to isolate the potential impact of estimated travel times on
acceptance rates.
Descriptive statistics are provided below for all variables in our model. Seventy-two (72) percent of the
sample were not deemed reduced-isolation students (Table 11). A little less than half the sample (42.9
percent) lived in Hartford (Table 12). The sample included students from a range of SES tiers, with lower-
income Tier A students representing 44 percent (Table 14). Table 15 shows the grade levels of the
schools to which families applied; a sizable amount applied to an early childhood school (27.9 percent).
Lastly, Table 16 displays the school rankings listed by parents on the lottery application.
Table GG shows the distribution of outcomes for first-round lottery offers. Nearly sixty percent (57.6
percent) of the offers were accepted and 18.4 percent were actively declined by parents. Another 24.0
percent were administratively declined by RSCO after not receiving a response from parents. For our
model, we collapsed active declines and administrative declines.
School ranking indicated by parents on the lottery application are summarized in Table HH. Parents can
rank up to five magnet schools and five Open Choice schools on their application. (A small percentage of
rankings (< 1 percent) in the data set were above 5, for an unknown reason.) For our multivariate model,
we reverse-coded these rankings so that a higher value reflected a stronger preference; this
transformation makes for easier interpretation of the results.
Table 11. Reduced-Isolation Student Status, 2022-23 Lottery
Reduced-Isolation
N
%
Not RI
8,045
72.0
RI
3,133
28.0
11,178
100.0
Table 12. Hartford Resident Status, 2022-23 Lottery
Hartford Resident
N
%
No
6,384
57.1
Yes
4,794
42.9
Total
11,178
100.0
44
Table 14. SES Tiers, 2022-23 Lottery
Tier
N
%
A (low)
4,925
44.1
B (med)
2,806
25.1
C (high)
3,447
30.0
11,178
100.0
Table 15. Grade Span of Applying School, 2022-23
School Grades
N
%
PK3-PK4
3,115
27.9
Elementary K-5
2,473
22.1
Middle School 6-8
2,263
20.2
High School 9-12
3,327
29.8
11,178
100.0
Table 16. School Rank on Lottery Application
Rank
N
%
1st
7,853
70.3
2nd
1,491
13.3
3rd
812
7.3
4th
548
4.9
5th
398
3.6
6th
46
0.4
7th
12
0.1
8th
9
0.1
9th
5
0.0
10th
4
0.0
45
Table 17. First-Round Lottery Placement Offer Outcomes for Magnets and Open Choice
Offer Outcome
N
%
Accepted
6,443
57.6
Admin Decline
2,680
24.0
Declined
2,055
18.4
Total
11,178
100.0
Figure 8 displays clustered bar charts that portray the relationship between driving time and parent
magnet offer decision. Driving times were collapsed into three categories. As emphasized above, these
are estimated driving times from home to school but are not estimated bus ride times. Notably, living
closer to an offered magnet school does not necessarily lead to higher acceptances relative to being
farther away. In fact, oddly, longer rides are more likely to be accepted. This analysis considers only two
variables and does not account for other likely influencers, such as choice ranking of school. (Later, we
incorporate other observables that may influence parent decision making in a multivariate model.)
Figure 8. Parent Decisions by Estimated Driving Time to School, 2022-23
Figure 9 shows the relationship between parent decisions on lottery offers to a magnet school and their
ranking of that school. Most (70.3 percent) first-round offers were to “first-choice” schools; however,
less than half those offers were accepted. One could read this figure as implying driving time has a slight
negative influence on accepting a seat. This is contrary to what one might expect.
46
Figure 9. Parent Decisions by Applicant School Ranking, 2022-23
Next, we turned to multivariate techniques to discern relationships among six predictor variables on
parent decision making.
The results of the binary logistic regression are summarized in Table 17. The columns labeled B and
Exp(B) in Table 17 help inform the contribution of each variable, on average, on the likelihood of an
offer acceptance. B represents the standardized regression coefficient for each variable, while holding
the remaining variables constant. Positive coefficients indicate a positive relationship with a decision to
accept; negative coefficients suggest the opposite influence. Generally, the larger the value of the
coefficient, the stronger its influence on the decision outcome. Exp(B) represents the odds ratio is the
predicted change in odds for a unit increase in the predictor. The “exp” refers to the exponential value
of B. When Exp(B) is less than 1, increasing values of the variable correspond to decreasing odds of the
event's occurrence. When Exp(B) is greater than 1, increasing values of the variable correspond to
increasing odds of the event's occurrence.
47
Table 17. Results of Binary Logistic Regression, Parent Decision to Accept a Magnet Offer as Influenced
by Five Predictor Variables
Variables in the Equation
B
S.E.
Wald
df
Sig.
Exp(B)
Step
1a
Driving Minutes Home to
Applying School
-.020
.003
53.676
1
<.001
.981
RI Student
-.367
.054
45.836
1
<.001
.693
Hartford resident
.347
.062
31.176
1
<.001
1.415
SES Tier
-.174
.035
24.601
1
<.001
.841
School Preference
.403
.020
403.992
1
<.001
1.497
Constant
-2.814
.215
171.963
1
<.001
.060
a. Variable(s) entered on step 1: Driving Minutes Home to Applying School, RI Student, Hartford resident, SES Tier,
School Preference.
The strongest predictor of accepting a lottery magnet seat, at least in this particular model, is school
preference.
19
We can convert the odds ratios listed in the last column (Exp(B)) to probabilities. Thus, for
every standard deviation unit increase in school preference score, the odds of accepting a magnet seat
increase by 49.7 percent. The next strongest influence on accepting is being a Hartford resident, which
increases the odds of accepting a magnet seat by 41.5 percent. The remaining variables exhibit a
negative influence on parents’ decision to accept a seat, with being a RI student having the strongest
relationship. Driving time to school has virtually no effect on parent decision making, after taking into
account the remaining variables in the model.
The five predictor variables explained parent decisions with a modest degree of accuracy. One way to
gauge model fit is to examine the classification table (Table 18). The table indicates that the model
accurately predicted “declines” 43.9 percent of the time, while it accurately predicted “accepts” at a
much higher rate of 81.6 percent of the time. The level of accuracy from our model is not surprising
given the range of influences on parent decision-making that were not captured by our five-variable
model, successfully predicting the decision outcome for 65.7 percent of cases.
19
This variable was reverse coded to make interpretation easier; originally, lower scores (e.g., 1, 2) reflected a
higher preference.
48
Table 18. Classification Table for Binary Logistic Regression
Classification Tablea
Observed
Predicted
Parent Decision
Percentage
Correct
Decline
Accept
Step 1
Parent
Decision
Decline
1739
2224
43.9
Accept
1005
4452
81.6
Overall Percentage
65.7
a. The cut value is .500
Limitations
Our analyses of bus routes, travel times, and lottery decisions in this section remains limited to learning
from families who either already participated or, in the case of the lottery, were interested in
participating in a Choice program. We did not capture families who did not apply to the lottery and do
not know the degree to which transportation played a role in their non-participation. In our logistic
regression analysis, we did not disentangle magnet offers from Open Choice offers.
Summary
In summary, our quantitative analysis generally aligned with the themes present in our interviews with
parents. Certain groups, such as Hartford students in the Open Choice program and suburban students
in interdistrict magnets, possibly face large burdens in terms of student transportation. For Hartford
students in Open Choice, total ride time and distance to school is especially high, possibly leading to
reduced opportunities in after-school activities and the like. For suburban students in interdistrict
magnets, the distance to their bus stops was especially far, echoing the concerns around bus stop
locations brought up by parents in interviews. In addition, the analysis suggests that where
transportation plays a factor in school choice decisions, it may largely occur either before the application
process, or once students are enrolled and experiencing transportation directly. The lottery data
suggests that once offered a placement, distance is a much smaller factor compared to personal ranking
of school in the application process. Regionality may matter to a large degree when students are
considering the schools they apply to in the first place, and so do not consider distance as heavily when
deciding to accept or reject placement offers.
49
Conclusions and Implications
Beyond the practical consequences, we consider the implications for equity and fairness for families.
Using a transportation equity lens, we learned that certain groups of students and their families,
especially those who are poor or minoritized or both, bear a greater burden in attending a school of
their choice. Long bus rides, multiple bus stops, limited afterschool late busing, and excessive burden on
families to drop off and pick up their child at the bus stop are among the main concerns.
RSCO commissioned this study to assess Choice program transportation, particularly from the
perspective of parents who are the primary decision makers with regard to student participation. Our
inquiries looked into possible disincentives associated with school transportation to participate in
Choiceparticipation in the form of accepting or declining a placement offer or leaving the program
once enrolled. We were cognizant of RSCO’s obligation to adequately meet parent demand for
attending an interdistrict school of choice and view the evidence from the perspective of its influence on
participation in the program, initial and continued.
In this section, we bring together findings from our multiple analyses in the form of meta themes. They
represent prominent, cross-cutting findings grounded in one or more of our data sources (i.e., interview
analysis, complaint log analysis, bus route and bus stop analysis, statistical modeling of lottery data). We
also discuss their implications, focusing on parent decision making in choice and transportation equity.
Transportation issues are inevitable in any large-scale busing enterprise. RSCO Transportation has made
concerted efforts to minimize travel disruptions and offers several mechanisms to assist parents with
travel. For instance, it offers an electronic notification system to inform parents via phone, email, or text
of any route delays and issue phone calls in case of a busing incident or accident. Although not
universally offered or used, a GPS tracking service is freely available for parents to locate their child’s
bus in real time. In the beginning of the school year, when the most transportation problems generally
occur, RSCO has front-loaded customer service representatives to meet the high load of consumer
inquiries. Other efforts include offering daily travel reimbursement under certain conditions for families
who drive their child to school. That said, a number of transportation concerns surfaced in our multi-
method analysis.
Bus Stop Locations and Commutes a Prominent Issue
Parents noted concerns with bus stops in the interview sample and complaint logs. The concerns
involved both the stop location themselves and the commute from home to the stop. Some parents
expressed concerns with bus stop venues, finding many of them unsafe due to high traffic, poor lighting,
or other dangerous conditions, all of which were compounded by long wait times. Parents also were
concerned with their child’s commute to the bus stops, citing long walks, unsafe routes (traffic, seedy
areas, darkness in the early morning or late afternoon, exposure to harsh weather conditions). For
many, the commute to the assigned stop was not walkable or deemed not walkable for their child, and
so they felt compelled to drive their child. This worked fine for some parents but presented a heavy
50
burden on others. Many noted it interfered with their work schedule or resulted in unexpected extra
expenses.
Personal Car Almost a Precondition
We got the sense that for some parents, a car was an absolute necessity to participate in a Choice
program. Of course, this was not a universal need, but a good number implied there was no other
option for them other than to drive their child to the bus stop or to the school itself. Usually it was some
combination of the bus schedule not fitting into their work schedule, dissatisfaction with the bus stop
location or commute to get there, accommodating a sibling’s transportation or child care needs, or
accommodating an afterschool activity. The convenience afforded by a car is consistent with other
research on school transportation (e.g., Lenhoff et al., 2023).
Distance to School only Minor Influence on Participation, but Significant Impact on
Equity
We learned that, generally, the distance to school did not hold a substantial influence on parent
participation in choice. A significant caveat is that our parent data included only those who had or were
already participating in choice either by enrolling their child in a choice school or applying to the lottery.
We found from most of analyses that school proximity from the home has limited influence on
participation. However, distance to school raised issues of inequity. Choice program parents and
students endured, on average, much longer commutes to school. For bused students, combining the
time to get from home to the bus stop with the bus ride to school added up substantially for some
students.
Inequitable Distances and Commutes to School
Hartford Open Choice students are faced with the longest distances from stop-to-school via our analysis
of expected bus ride travel times. When factoring in that many suburban students are likely driven to
their bus stops, Open Choice students likely face the longest total home-to-school travel as well.
Hartford magnet students have shorter commute time to school than Open Choice students from
Hartford. This is not particularly surprising given the geographic spread of Open Choice schools;
however, the choice zones have been an attempt to reduce those distances and commute times and
Open Choice students from Hartford still have longer commutes.
Suburban magnet students are traveling farther to magnets, on average, than Hartford magnet
students; we think this is at least partly due to the need to travel through congested or slower traffic
areas in Hartford. For instance, suburban students in CREC magnet schools have the longest time
relative to Hartford Host magnets. Suburban CREC magnet students stand out with a median expected
ride time of 41 minutes and 20.3 percent of those students have a ride of greater than 60 minutes. CREC
has only 7 magnets in central Hartford and with many more situated in the second ring. This may be
because regionality plays less of a role in recruitment and applications for these schools than for
Hartford Host magnets. Some Hartford Host magnets, particularly on the edges of the Hartford Metro,
appear more likely to have lower estimated median driving times, supporting the idea that they may
51
exert a higher local draw compared to CREC magnets in the second ring who may draw students from
farther away.
We witnessed some effects of distance of school on parent participation in choice, but likely the
influences are already baked into decisions to enter lottery in the first place and list schools that are
feasible for them. Ultimately, though the realities of transportation and bus stops aren’t realized until
enrolled in the program. We see an exception of distance mattering most when we examined from the
perspective of sending district, again supporting the idea of regional draw. Distance, unsurprisingly, may
matter most for students who are farthest away from metro Hartford for whom bus rides would be
exceptionally longer, and personal adaptations (i.e. driving to school, carpooling, responding to delays or
changes in schedule) would be more complicated.
Based on our analyses, distance to schools was not a major barrier to participation in choice. Our data
did not, however, represent families who were making decisions about applying for a school a choice.
Many non-participants undoubtedly eschew school choice because of their preference to attend schools
close to home. Further, we suspect lottery parents considered travel distance prior to applying. We did,
however, find significant disparities in ride times to school between students attending intra-district
magnet schools and their local peers in traditional public schools. While to some degree this is to be
expected, it is certainly a factor that may impact decisions for some families.
Economic Accessibility
Our analyses suggest that, because of transportation [concerns], participation in choice was at least in
part a function of household resources. For school choice to be truly accessible to all families,
transportation factors should not prevent them from enrolling their child in a desired Choice school. But
based on our interview data, it seems that it did. At the worst, parents pulled their child from the Choice
program. Other parents remained in Choice but were unfairly burdened for their participation. We
heard from several parents who used personal transportation or who otherwise made personal
sacrifices to get their child to school. Either by preference or perceived need, many of these parents
drove their child to the bus stop or school. Only those parents who had the baseline means to do so
could afford the time and resources required to do so. Suburban families, in general, are economically
better off than Hartford families. They also have fewer single-parent households. Suburban families are
more apt to be able afford childcare, be at bus stops waiting for bus with child, and have personal
transportation, and finally, can be more adaptive to unforeseen circumstances and can offer time
flexible extra as opposed to relying on school transportation. Some parents, particularly stay-at-home
moms, have the freedom to drive their child and access to a car. Because of this, one could argue Choice
is more accessible to them relative to families with fewer resources.
In their study of families in choice-rich Detroit, Lenhoff et al. (2023) note:
The second most common daily mode to school after your own car was “multiple modes,” when
parents indicated more than one daily mode to school. Students who used multiple modes may
be uniquely disadvantaged. They had the lowest average family income of all transit modes at
$20,640, and they had among the lowest car ownership at 51%. This suggests that families
52
without cars may be coordinating several different transportation modes for their children as
they can find them. (p. 13).
53
Recommendations & Ideas for
Consideration
In this section we offer recommendations for policy and practice. Our recommendations are based on
our quantitative and qualitative analyses, and in some cases are supported by additional information,
including our reviews of RSCO Transportation materials and the school transportation research
literature. Our intention is to offer ideas for program improvement and future areas of discussion. Policy
makers should continue to look for ways to improve upon transportation from the perspective of
parents and students. The transportation ecosystem is more than just bus routes and schedules; it
involves the entirety of how parents may experience it, from wake up to the return home. Even small
changes may mean a lot to students and their parents, and possibly increase the likelihood of
participating and persisting in the school choice program. Finally, we encourage policy makers to
examine transportation through a mobility justice framework, which considers how transportation may
be experienced differently by different populations (Bierbaum et al., 2021). Achieving equity for all is not
just for equity’s sake, doing so will also open accessibility to choice.
Recommendations
Improve Conditions for Getting Students from Home to the Bus Stop
A major complaint among parents was the home-to-bus stop experience. We suggest RSCO target
improvements in the neighborhood and centralized bus stops. Parents desired stops that were closer
and more convenient. It may be worth exploring possible improvements; for example, increasing the
number of bus stops or implementing a shuttle service. For instance, a survey of Detroit families
reported that 74 percent of parents indicated that it would be “very helpful” or “helpful” to have bus
pick-up at their house, and 56 percent indicated it would be “very helpful” or “helpful” to have a pick-up
stop within 0.25 miles of their house (Lenhoff et al., 2023). There is some research that suggests busing
students who are normally required to walk to school reduces absenteeism (Sattin-Bajaj, 2018).
Ensure All Stop Locations Are Safe
There was some concern about the safety of some stop locations. Parents expressed multiple concerns
about stop safety. These concerns ranged from busy streets without sidewalks, dark or poorly lit areas, a
lack of shelter from weather, all the way to possible crime or violence at the stop location itself. We
suggest RSCO strategically place stops, particularly centralized stops that may be a distance away from
students’ neighborhoods, in locations with sidewalks, active lighting, public infrastructure, and other
features that increase the safety of students waiting at or traveling to them. In addition, we encourage
RSCO to continue to be responsive to parent concerns about specific locations and act quickly to re-
evaluate in the case of safety concerns.
54
Recalibrate the Complaint Type Categories in the RSCO Online Complaint Form
Parent transportation complaints are recorded and stored in a RSCO database, providing the
opportunity to periodically assess service quality and to identify areas for improvement. Some of the 12
categories potentially overlap or are broader in scope than others. The fairly commonly applied “Other”
category yields little information as framed. An analysis of the description of the issue listed alongside
each complaint category suggests some mis-categorization based on the current categories. For
instance, “Bus/Vehicle Driver” often contained descriptions of issues that went beyond the actual driver.
The clearer the recorded complaint type, the more accurate the feedback provided to the service
contractor.
Make the Online Complaint Form More Prominent on the Website
RSCO Transportation offers several channels for parents to submit a transportation complaint. Most
complaints are issued via phone calls to the customer service office. The online complaint form could be
made more accessible on the main page of the website, perhaps next to the customer service phone
number (https://www.crec.org/transportation/rsco.php). As it stands, parents would have to know that
the form exists under the general “Forms” link in the side menu.
Revisit Bus Notification System to Improve Upon Efficiency
The notification system does not always appear dependable or timely for families. Perhaps this is due to
some inherent delays in the process; for instance, a bus driver must first make a determination that they
will be running late, then notify a central dispatcher who may also be tied up taking calls from other bus
drivers or even families, and the dispatcher then presumably sends a blanket notification to all families
on the route in question.
Ensure Families have Access to All Bus Notification Mechanisms
RSCO Transportation offers notification services to families when buses are running late or there are
incidences such as a vehicle breakdown. Phone calls are made in those instances and families also
appear to have the option to receive text and emails if they make such arrangements with CREC
Transportation. In addition, a GPS tracking app is also available for First Student and DATTCO families to
monitor the location of their child’s bus in real time. A third bus provider, Transportation Management
Services, is intending to provide a GPS service (currently listed as “coming soon” on the RSCO
Transportation website). Because this seems like an invaluable resource for families, it may be worth
gathering data on the extent to which families are accessing and using these services; there may remain
unknown impediments to their access and proper use.
Involve Families and Students in Developing Transportation Policy
Involve parents and students actively in shaping transportation policies. Beyond simply hearing from
those most impacted by school transportation, engaging parents and students genuinely in policy
development is prudent. Policy makers could “look to the creative resources families use to solve their
transportation problems for potential answers” (Lenhoff et al., p. 356). They could also be “active
participants in designing information about the transportation resources available and ensuring that
their peers have a deep understanding as they make school enrollment decisions” (p. 356).
55
Look to Other Models for Innovative Approaches
Explore travel innovations attempted in other cities with rich histories of school choice (see, for
example, Vincent et al. (2014). For instance, Denver…. “Additionally, the Denver Regional Council of
Governments (DRCOG) provides schools and families with access to a “SchoolPool” program. SchoolPool
matches families at a school or nearby schools based on the proximity of household residences. After
being matched, families can organize carpools, biking or walking groups, or group travel via public
transit in order to get to school. Nearly 70 schools across the greater Denver area actively participate in
the SchoolPool program. In the 2013–14 school year, SchoolPool provided over 15,000 family matches” (
https://waytogo.org/for-commuters/schoolpool). SchoolPool also used in Charlottesville (Vincent et al.,
2014).
Communicate Transportation Options to Prospective Choice Parents
Parents making decisions about which schools to apply to, and how to rank them, may benefit from
specific information on travel time and bus stops. Parents undoubtedly factor distance to school as they
consider schools of choice. They may be less likely to know about bus transit schedules and travel times,
or what options are available to them before and after school.
Ideas for Consideration
Walking Chaperones
For younger students who have a long walk to a bus stop or who live too close to their school to be
bused, adding a chaperone to walk with students may help reduce parent fears about children walking
in high traffic areas unaccompanied by an adult. It may be worth exploring if there are opportunities for
groups of students to walk together with a chaperone. The nature of Greater Hartford School Choice,
with students at any one magnet school coming from widely dispersed areas, may not lend itself to this
option. If, however, clusters of students are within walking distance of either their school or a bus stop,
it might be worth looking at the Walking School Bus model
20
(National Center for Safe Routes to School,
2006) or less formal neighborhood-initiated programs. Kang and Diao (2022) found that walking school
bus programs are feasible even in a low-density suburban setting.
Public Transit Passes
For older students, public transit may present a feasible alternative to busing or driving to school. Of
course, this is not a universally viable option for students, as the transit route has to align with the
student’s path to school. Parents and students have to be comfortable doing so, as well. If this were at
all a useful option for students, RSCO could help negotiate free or discounted passes. This practice has
20
See http://guide.saferoutesinfo.org/walking_school_bus/
56
been shown to be difficult to implement in some cases, however, and the research is mixed on
successes for students (e.g., Fan & Das, 2016; McDonald et al., 2004). Wexler et al. (2021) found that the
Minneapolis Go-To Student Pass Program substantially reduced excused absences among participating
students.
Further Regionalization of Choice Programming
Without substantively reducing choices, regionalizing school choice even more so than is done now
could lead to shorter travel times, more clustering of students to transport, and overall greater
efficiencies in transportation. The evolution of the interdistrict magnet schools in the Sheff region over
the past two decades has resulted in a substantial number of new schools being built. They span various
grade levels, locations, and curricular themes. For instance, interdistrict magnets include twelve
different grade configurations. One could view their evolution as haphazard, to a degree. Open Choice
offers options across metro-Hartford and into the outer ring and exurbs. RSCO introduced four
geographical zones in Hartford to help regionalize choice sets for families; this undoubtedly assisted
transportation to be more efficient and deliver shorter commutes. More recently, RSCO established,
where possible, magnet school pathways so that students could envision their entire PK3-12 journey;
pathways also give magnet students some assurance that when they grade out of an elementary
magnet, for instance, it isn’t left to pure chance that they could enroll in an upper-grade magnet. Given
the number of magnet schools, RSCO could explore the potential for clustering or grouping them (akin
to the zoning concept). It could not only promote efficiencies in areas like transportation, it could create
more of a neighborhood-type school feeling for students and families. Specific schools could also recruit
in neighborhoods, for instance. Like a clustering approach. It might help achieve Sheff goals if you target
neighborhoods that are Tier C? The only concern is the spread of possible neighborhoods in the suburbs;
but I guess that’s also the point, to try and gain some efficiencies. Could look for a proof of concept here
by targeting schools that are way out of compliance in terms of RI and try intentionally recruiting in
batch in those census blocks that help meet the RI goals (e.g., target Tier C); it is akin to reverse-
engineering.
Revisit the 20-minute Wait Time Window
The expectation that students/families should arrive at a bus stop 10 minutes earlier and potentially
wait 10 minutes after the scheduled pickup time is prominently noted in the RSCO Transportation Family
Information Handbook. The policy also indicates “The bus does not have to wait until the scheduled
pickup time before leaving the bus stop” (p. 12). To be sure, these guidelines may be standard operating
procedure for bus transportation writ large. It did raise the question whether families and students are
fully aware of these guidelines and also how much they come into play. Most parents and students are
savvy enough to make adjustments to the actual bus schedule; therefore, the 20-minute window may
not be a major issue. Nevertheless, we suggest examining early/late pick-up and drop-off records for all
bus stops over an adequate sampling period to assess how often a 20-minute window is needed, how
often buses do arrive early, and which tend to arrive consistently early or late. At the very least make
sure parents have received this message about what constitutes “on time” status for buses.
57
Suggestions for Future Research
The findings from this study offer a point of departure for additional targeted research to learn more
about the RSCO transportation ecosystem.
Using a survey or focus groups, gather feedback from Choice students on how they experience
the commute to and from school. Speaking to students may reveal unforeseen benefits and
drawbacks of various modes of transportation. They could also speak to the morning and
afternoon routines and how those may affect sleep or activities after school, school-sponsored
or otherwise. Does their ride hinder or facilitate completing homework? For those who ride the
bus, what is their experience like?
Survey parents who participate in a Choice program to reach a larger sample or a targeted
sample based on the findings here. Several distinct themes emerged from our interview analysis
that would inform survey generation. For instance, gather data on parent mode(s) to school.
Often times in survey research it is wise to first conduct open-ended queries through interviews
and observations to inform subsequent surveys, which are dominated by closed-ended items. In
other words, it is difficult to anticipate what topics to ask parents about without getting a better
sense of what they are through open-ended techniques. Surveys are typically dominated by
closed-ended items.
Survey a sample of parents who have not participated in the choice lottery. Doing so would
capture sentiments of families who have opted not apply to determine the extent to which
transportation factors play a role.
[1] See http://guide.saferoutesinfo.org/walking_school_bus/
58
References
Bierbaum, A. H., Karner, A., & Barajas, J. M. (2021). Toward mobility justice: Linking transportation and
education equity in the context of school choice. Journal of the American Planning Association, 87(2),
197-210.
Burgess, S., Greaves, E., Vignoles, A., & Wilson, D. (2015). What parents want: School preferences and
school choice. The Economic Journal, 125(587), 1262-1289.
Chubb , J. E., & Moe T. M. (1990). Politics, markets, and America’s schools. Brookings Institution.
Cobb, C. D., Connery, C., & Wentzell, C. (2021, December). Magnet school student attrition: An
exploratory analysis of why students leave. Report Prepared for the Regional School Choice Office,
Connecticut Department of Education.
Cobb, C. D., & Connery, C. (2021, March). School lottery family decision analysis: Patterns among
decliners (Final Report). Submitted to Connecticut Department of Education, Regional School Choice
Office.
Corcoran, S. P. (2018). School choice and commuting. The Urban Institute.
Cordes, S. A., Leardo, M., Rick, C., & Schwartz, A. (2019). Can school buses drive down (chronic)
absenteeism. In M. Gottfried & E. Hutt (Eds.), Absent from school: Understanding and addressing
student absenteeism (pp. 121136). Harvard Education Press.
Cordes, S. A., Rick, C., & Schwartz, A. E. (2022). Do long bus rides drive down academic outcomes?
Educational Evaluation and Policy Analysis, 44(4), 689-716.
Edwards, D. S. (2021). Just out of reach? Unrestrained supply, constrained demand, and access to
effective schools in and around Detroit. Educational Evaluation and Policy Analysis, 43(3), 391-418.
Fan, Y., & Das, K. V. (2016). Assessing the impacts of student transportation via transit on student
attendance and academic achievement. In Transportation Research Board 95th Annual Meeting,
Washington, DC.
Hastings, J. S., Kane, T., J., & Staiger, D. O. (2005). Parental preferences and school
competition: Evidence from a public school choice program. National Bureau of
Economic Research Working Paper No. 11805. National Bureau of Economic
Research, Cambridge.
He, S. Y. (2013). Will you escort your child to school? The effect of spatial and temporal constraints of
parental employment. Applied Geography, 42, 116123.
59
He, S. Y., & Giuliano, G. (2018). School choice: Understanding the trade-off between travel distance and
school quality. Transportation, 45(5), 14751498.
Kang, B., & Diao, C. (2022). Walking school bus program feasibility in a suburban setting. Journal of
Planning Education and Research, 42(3), 365-374.
Lenhoff, S. W., Singer, J., Stokes, K., Mahowald, J. B., & Khawaja, S. A. H. A. R. (2022). Beyond the bus:
Reconceptualizing school transportation for mobility justice. Harvard Educational Review, 92(3), 336-
360.
Lenhoff, S. W., Singer, J., Brown, D., & Pogodzinski, B. (2023, March 23-25). School transportation mode
and student attendance across schools of choice [Paper presentation]. Association for Education Finance
and Policy (AEFP) Annual Conference, Denver, CO.
Lidbe, A., Li, X., Adanu, E. K., Nambisan, S., & Jones, S. (2020). Exploratory analysis of recent trends in
school travel mode choices in the US. Transportation Research Interdisciplinary Perspectives, 6, 100146.
Lincove, J. A., & Valant, J. (2018). New Orleans students’ commute times by car, public transit, and school
bus. Urban Institute.
Makarewicz, C. (2013). Vouchers, magnet schools, charter schools, and options: Analyzing the effects of
school and housing choices on mode choice to school. Transportation Research Record, 2327(1), 1-8.
Mandic, S., Sandretto, S., Bengoechea, E. G., Hopkins, D., Moore, A., Rodda, J., & Wilson, G. (2017).
Enrolling in the closest school or not? Implications of school choice decisions for active transport to
school. Journal of Transport & Health, 6, 347-357.
McDonald, N. C., Librera, S., & Deakin, E. (2004). Free transit for low-income youth: Experience in San
Francisco Bay Area, California. Transportation Research Record: Journal of the Transportation Research
Board, 1887(1), 153160.
National Center for Safe Routes to School (2006). The walking school bus: combining safety, fun and the
walk to school. Chapel Hill, NC.
Orfield, G., & Frankenberg, E. (2013). Educational delusions? Why choice can deepen inequality.
University of California Press.
Pattillo, M. (2015). Everyday politics of school choice in the Black community. Du Bois Review: Social
Science Research on Race, 12(1), 41-71.
Phillips, R., Hausman, C., & Larsen, E. S. (2012). Students who choose and the schools they leave:
Examining participation in intradistrict transfers. The Sociological Quarterly, 53(2), 264294.
Rhoulac, T. D. (2005). Bus or car? The classic choice in school transportation. Transportation Research
Record, 1922(1), 98104.
60
Sattin-Bajaj, C. (2018). “It’s hard to separate choice from transportation’’: Perspectives on student
transportation policy from three choice-rich school districts. Urban Institute.
Sattin-Bajaj, C. (2023). Student transportation in choice-rich districts: Implementation challenges and
responses. Education Finance and Policy, 18(2), 351-364.
Sattin-Bajaj, C., & Roda, A. (2020). Opportunity hoarding in school choice contexts: The role of policy
design in promoting middle-class parents’ exclusionary behaviors. Educational Policy, 34(7), 992-1035.
Scott, M. R., & Marshall, D. T. (2019). Public transit and school choice in Philadelphia: Exploring spatial
equity and social exclusion. Journal of School Choice, 13(2), 177197.
Seidman, I. (2013). Interviewing as qualitative research: A guide for researchers in education &
the social sciences (4th ed.). Teachers College Press.
Sheff v. O'Neill (1996). 238 Conn. 1, 687 A.2d 1267.
Trajkovski, S., Zabel, J., & Schwartz, A. E. (2021). Do school buses make school choice work? Regional
Science and Urban Economics, 86, 103607.
Teske, P., Fitzpatrick, J., & O’Brien, T., (2009). Drivers of choice: Parents, transportation,
and school choice. Seattle. Center for Reinventing Public Education, University of
Washington, Bothell.
Valant, J., & Lincove, J. A. (2023). Transportation inequities and school choice: How car, public transit,
and school bus access affect families’ options. Educational Researcher, 52(9), 535-543.
Vincent, J. M., Makarewicz, C., Miller, R., Ehrman, J., & McKoy, D. L. (2014). Beyond the yellow bus:
Promising practices for maximizing access to opportunity through innovations in student transportation.
Center for Cities & Schools.
Weinstein, M., Cordes, S. A., Rick, C., & Schwartz, A. E. (2022). Riding the yellow school bus: Equity in bus
transportation across districts, schools, and students. Urban Education, 00420859221114084.
https://doi.org/10.1177/00420859221114084
Wexler, N., Ryan, G., Das, K., & Fan, Y. (2021). Free transit passes and school attendance among high
school students. Transportation Research Record, 2675(8), 135-147.
Wilson, E. J., Marshall, J., Wilson, R., & Krizek, K. J. (2010). By foot, bus or car: Children’s school travel
and school choice policy. Environment and Planning A: Economy and Space, 42(9), 2168-2185.
Yettick, H. (2016). Information is bliss: Information use by school choice participants in Denver. Urban
Education, 51(8), 859-890.
61
Appendix A
Semi-Structured Interview Protocol with Prompts - Parent and Caregiver
1. What is your relationship to the child/ren?
2. What grade is/are the child(ren) in? Which school do they attend?
3. How many years has the child attended this school?
4. How does your child/ren get to school?
5. Describe your morning routine on a school day. Some questions to consider:
a. What time do you wake up?
b. How much time does the bus take to get to your home?
c. What time does it arrive?
d. How long do you wait at the stop?
e. Where do you go afterwards?
6. What happens when your child misses the bus?
7. Describe the bus stop experience.
8. What has your child said about their busing experience?
9. Has your child been to a different school, if so what was the busing experience?
10. If yes, How has the busing experience been different between this school/previous schools or
this year/previous years?
11. How was transportation a factor in your decision making process around school choice?
12. Describe your typical return from school process?
13. What are some abnormal days like?
a. Missed bus
b. Snow/weather
c. Late Bus
14. What else do you think I/we should know about your experience?
15. Knowing what you do now about school transportation, if you were to redo your school choice
process, would transportation impact your decision?
62
Appendix B
Figure B1. Estimated Commute Time Between Home and Open Choice Seat Offer for Hartford Residents
by Receiving District (20 or More Students)
63
Figure B2. Estimated Commute Time Between Home and Magnet School Seat Offer for Hartford and
Suburban Students by Sending District (30 or More Students)
The median estimated travel time for all students offered a magnet seat was 15 minutes. When broken
down by magnet school, we observed considerable variation around that overall median. Multiple
schools had an estimated median of over 20 minutes, including Reggio Magnet, which produced the
highest frequency of complaints in the complaint analysis. The highest estimated median travel was
Global Experience Magnet School with 24 minutes (Table 7). 86.5 percent of students had an estimated
travel time of 15 or more minutes, while 26.4 percent had estimates of 30 or more minutes, which was
the highest percentage for both lengths of estimated travel time relative to any other school. On the
other end of the spectrum, the Montessori Magnet at Batchelder had the lowest median estimated
travel time of 9 minutes, but did not have the lowest percentage of students with expected travel of
either greater than or equal to 15 or 30 minutes (32.3 percent and 5.3 percent, respectively). Of schools
with N >100, Connecticut River Academy at Goodwin had the lowest percentage of students with
estimated travel of 15 minutes or more with 22.2 percent and a median of 10, while Wintonbury Early
Childhood Magnet School had the lowest percentage of students with estimated travel of 30 minutes or
more with 0 percent, and a median of 15 minutes. To some degree, these differences on the lower end
of the estimated travel spectrum may indicate that regionality or distance might have a stronger impact
on the pool of applicants for some schools than others.
64
Table B1. Estimated Travel Time Between Student Home and Offered Magnet School Placement by
Offering School
Offering Magnet School
Median
Estimated
Drive Time
(min)
Percent
≥ 15 min
Percent
≥ 30 min
N (Placement offers
Academy of Aerospace and Engineering
19
83.3
19.0
221
Academy of Aerospace and Engineering
Elementary
17
74.4
11.9
320
Academy of Computer Science and
Engineering
22
87.1
13.5
325
Academy of Computer Science and
Engineering MS
11
32.4
3.0
531
Academy of International Studies
20
78.0
12.8
327
Academy of International Studies
Elementary
14
49.3
2.3
302
Academy of Science and Innovation
19
68.2
7.4
258
Ana Grace Academy of the Arts
23
85.6
22.1
208
Betances Learning Lab
11
37.6
2.1
330
Betances STEM Magnet School
13
47.2
7.6
144
Breakthrough Magnet School, North
15
54.3
4.0
173
Breakthrough Magnet School, South
12
37.9
2.4
124
Capital Preparatory Magnet School
13
40.8
3.6
331
Classical Magnet School
12
50.3
2.4
165
Connecticut IB Academy
17
67.9
6.4
109
Connecticut River Academy at Goodwin
10
22.2
2.3
311
Discovery Academy
11
29.2
5.2
154
Early College Advanced Manufacturing
10
22.2
0.0
18
Environmental Sciences Magnet at
Hooker
15
25.6
2.4
246
Glastonbury/East Hartford Magnet
School
16
63.6
2.8
316
Global Experience Magnet School
24
86.5
26.4
178
Great Path Academy at MCC
15
38.2
3.6
165
Greater Hartford Academy of Arts (HD)
19
71.1
14.0
342
Greater Hartford Academy of the Arts
14
49.3
10.9
138
Greater Hartford Academy of the Arts
High School
17
55.3
14.2
190
Hartford Magnet Trinity College Academy
13
43.2
6.4
250
Hartford PreKindergarten Magnet School
12
37.1
1.0
308
Kinsella Magnet School of Performing
13
43.5
7.7
299
Montessori Magnet at Batchelder
9
32.3
5.3
189
Montessori Magnet at Fisher
15
50.3
5.0
161
Montessori Magnet School (CREC)
16
53.1
4.0
177
Museum Academy
14
48.5
3.0
198
Pathways Academy of Technology and
Design
11
30.6
1.9
157
Reggio Magnet School of the Arts
20
80.8
12.1
240
Riverside Magnet School at Goodwin
11
26.4
0.8
239
65
Sport and Medical Sciences Academy
14
45.9
6.8
207
STEM Magnet at Annie Fisher School
15
54.7
5.0
179
University High School of Science and
Engineering
16
54.4
5.1
158
University of Hartford Magnet School
14
49.8
6.3
255
Webster Micro Society Magnet School
10
50.0
7.7
246
Wintonbury Early Childhood Magnet
School
15
51.3
0.0
232
ResearchGate has not been able to resolve any citations for this publication.
Technical Report
Full-text available
This report presents findings from an analysis of attrition of students enrolled in interdisciplinary magnet schools within the Sheff region. There were two main strands of analysis. The first was a qualitative study of exit interviews from a sample of families with children who were enrolled in a Sheff region magnet school at any point between the Fall of 2017 and the Spring of 2020, but voluntarily terminated their child’s enrollment to attend another school. The rationale for terminating their child(ren)’s enrollment was discussed and the larger factors that influenced their decision making were noted. The second strand was a quantitative analysis based on secondary data provided to us by the Connecticut State Department of Education (CSDE) as well as publicly available school-level data from the 39 facilities in the region. This portion of the analysis focused on whether students in non-transitional grades attending one of the schools returned to the same facility the next school year.
Technical Report
Full-text available
This report presents findings from an analysis of Connecticut's Regional School Choice Office (RSCO) lottery. There were two main strands of analysis. The first was a quantitative analysis of the administrative data from the 2019-2020 RSCO lottery. The goal of this analysis was to explore patterns among lottery acceptances, declines, and administrative declines of offers for interdistrict magnet schools and Open Choice schools. Specifically, lottery offer outcome data were disaggregated by relevant demographic subgroups, geographies, offer rounds, applicant choice preferences, and choice programs. We looked for associations between lottery offer outcomes and various moderating variables that may explain those outcomes. The second strand was a qualitative analysis of data collected from interviews with families who – actively or passively – declined offers from the RSCO Lottery. The quantitative portion of the study allowed us to explore broader patterns behind what patterns exist among lottery acceptances, declines, and administrative declines. The qualitative inquiry complements this information by seeking understand why families actively or passively declined offers from the school choice lottery, as well as how families are experiencing the lottery process and what, if any, aspects of the lottery process may help explain their declining.
Article
This study examines how the student transportation options available to families affect which schools are accessible to them in a choice-based setting. The study has two parts. First, we compare commute times by foot, public transit, school bus, and car. We show that providing school bus service reduces commute times and improves access for families without cars, but access to a car fundamentally shapes families’ options. Second, we explore the relationship between neighborhood-level measures of vehicle access and families’ school requests and placements. Car access is strongly associated with school requests and placements even after accounting for neighborhood characteristics. We consider car access as a pathway by which wealth disparities produce educational disparities in settings that emphasize school choice.
Article
This essay combines an ecological perspective with a mobility justice theoretical framework to reconceptualize the relationship between school transportation and educational access. Authors Sarah Winchell Lenhoff, Jeremy Singer, Kimberly Stokes, James Bear Mahowald, and Sahar Khawaja document the problem of “getting to school” that is at the intersection of students’ family, community, and social contexts and how it goes beyond whether there is a reliable mode of physical transportation. Bringing together a historical analysis of the policy landscape and interview data from parents and students in Detroit, they find that school transportation problems reflect the unequal political, social, and economic context in which families navigate enrollment and attendance. They discuss how policy makers can advance mobility justice in school policy by equitably distributing transportation resources, engaging students and parents as experts in developing and communicating transportation policy, and using institutional power to remedy structural barriers to educational access.
Article
While the yellow school bus is emblematic of American education, there is variation in bus service across the country. Millions of students rely on the bus, making transportation an essential part of a student’s educational experience. Thus, equity in transportation access is a key component of educational equity. Using national district-level data and New York City student-level data, we examine equity in transportation finance and bus provision. Our findings indicate significant racial and ethnic disparities, both nationally and in NYC. These results may help schools and districts design bus services to level the playing field and enhance student outcomes.
Article
Despite a growing recognition of the significance of student transportation for promoting equitable school choice, to date, there has been limited understanding of the implementation of school transportation policies, particularly in choice-heavy settings. Moreover, little is known about the challenges associated with managing student transportation in large school districts or how school- and district administrators working in distinct contexts respond to these challenges. This brief draws on interview data collected from district administrators, charter school leaders, charter authorizers, and stakeholders in three choice-rich school districts with distinct transportation policies and public transit infrastructures—Detroit, New Orleans, and New York City—to identify the common challenges associated with implementing student transportation policies and administrators' responses to them. The brief includes a discussion of the implications of high transportation costs, logistical complexity, and student safety challenges as well as non-uniform transportation provision and transportation policies for equitable access to charter schools. It ends with a review of some potential policy alternatives that policymakers should consider when designing transportation policies or working to improve existing transportation practices.
Article
Research concerning family preferences for schooling indicates that they value proximity to home as much as academic quality when choosing schools. However, preferences for proximity likely represent inability to access schools farther away from home, especially for disadvantaged students. I test whether distance and district boundaries constrain access to high-performing and effective schools for Detroit students where families choose between intradistrict, interdistrict, and charter schools, as well as an assigned school. I employ a unique data set that includes enrollment records, addresses, and commute times for Detroit residents regardless of where they attend school. Results show that disadvantaged students have little access to the highest quality schools available, specifically those outside Detroit. However, students attend higher performing schools within Detroit.
Article
In multiple U.S. cities, school districts and transit providers collaborate to provide students free public transit access, either replacing or supplementing the yellow school bus service. Although practitioners widely acknowledge these programs as a promising and innovative solution to absenteeism, there is no empirical research evidence to date confirming the benefits of these programs to individual students’ school attendance. We respond to this research gap by investigating the impacts of the Minneapolis Go-To Student Pass Program—a transportation program that began to provide student access to public transit in August 2013—on individual student school attendance. Using Poisson regression, we estimate both the traditional Difference-in-differences models and Two-Way Fixed Effects models to quantify effects of pass use and pass eligibility on student attendance. We find that student-reported pass use and pass eligibility reduce excused absences by 11.5% and 27.5%, respectively. Restricting our sample to students living in the School Walk Zone—the area within 2 mi of one’s school—we find that these effects are even more substantial, with pass use and pass eligibility reducing excused absences by 30.5% and 37.6% respectively. These findings imply that providing free access to public transit is broadly useful to improve student attendance, even for students who live within walking distance of the school and may not use transit passes regularly.
Article
While school choice has been well studied, there is little existing research exploring the role of transportation, in general, and school buses, in particular, to school choice decisions. We examine the effect of school buses on school choice decisions using data on kindergarten students and their eligibility for transportation assistance in New York City public schools in 2017. Using both conditional logit school choice models and regression discontinuity designs, we provide both descriptive and credibly causal evidence on the impact of school proximity, bus eligibility, and their interaction on school choice decisions. Our results indicate that proximity and buses both matter. Specifically, while distance significantly deters choice, school bus eligibility increases the likelihood of choosing a school by 1.4 to 4 percentage points (or 12 to 30 percent). Compared to a high-quality school, we find that bus eligibility has twice as large an impact on reducing the negative distance effect in the 0.5 to 1 mile range from school (27% versus 12%). These results will be useful for policy makers looking to leverage school transportation policy to improve school choice decisions, and ultimately student outcomes.
Article
Problem, research strategy, and findings Transportation equity research addresses questions of participation and planning process or the distribution of transportation’s benefits and burdens. This work largely elides issues related to youth and student travel. Existing work on school trip mode choice does not engage deeply with the equity implications of its findings. Education equity researchers identify the abstract notion of choice as equity enhancing but do not engage with the realities of disparate transportation resources and infrastructure. In this review essay, we articulate the contours of transportation equity, youth travel, and school choice research. We use emerging perspectives on “mobility justice” to frame the issues and provide vocabulary that can help bring education and transportation together in planners’ everyday work. Takeaway for practice A mobility justice framework encourages critical thought and action to address the root causes of inequities. Our conclusions make three contributions to research and practice: 1) challenging school district leaders to think about choice systems designed for access to schools, not just access to information about available options; 2) clarifying the broader implications of school choice by refusing to look away from the racial implications of forecasts; and 3) elucidating the need for a regional perspective.