K12EDUCATION
Departmentof
EducationShould
ProvideInformation
onEquityandSafety
inSchoolDress
Codes
Accessible Version
ReporttoCongressionalAddressees
October 2022
GAO-23-105348
United States Government Accountability Office
United States Government Accountability Office
GAOHighlights
Highlights of GAO-23-105348, a report to
congressional addressees
October 2022
K-12 EDUCATION
Department of Education Should Provide Information
on Equity and Safety in School Dress Codes
What GAO Found
While school districts often cite safety as the reason for having a dress code,
many dress codes include elements that may make the school environment less
equitable and safe for students. For example, an estimated 60 percent of dress
codes have rules involving measuring students’ bodies and clothing—which may
involve adults touching students. Consequently, students, particularly girls, may
feel less safe at school, according to a range of stakeholders GAO interviewed.
According to GAO’s nationally generalizable review of public school dress codes,
districts more frequently restrict items typically worn by girls—such as skirts, tank
tops, and leggings—than those typically worn by boys—such as muscle shirts.
Most dress codes also contain rules about students’ hair, hair styles, and head
coverings, which may disproportionately impact Black students and those of
certain religions and cultures, according to researchers and district officials.
Department of Education (Education) officials told GAO they are considering
options to provide helpful resources to stakeholders and the public, but as of
September 2022, Education had not provided information on dress codes.
Providing such information would align with the agency’s goal to enhance equity
and safety in schools.
Items Commonly Prohibited by School Dress Codes
Accessible Data for Items Commonly Prohibited by School Dress Codes
Percentage of school districts with dress codes that prohibit at least one clothing
item typically worn by...
Category
Lower bound
Estimate
Upper bound
Boys
62.2
68.8
75.5
Girls
85.3
90.3
94
Example boy items: Muscle shirts; Sagging pants
Example girl items: Midriff-bearing tops/spaghetti straps; Short skirts; Leggings
as pants
Source: GAO review of school dress codes; stock.adobe.com (base artwork). | GAO-23-105348
Schools that report enforcing strict dress codes predominantly enroll Black and
Hispanic students and are more likely to remove students from class. GAO’s
analysis of national data found that more than four in five predominantly Black
schools and nearly two-thirds of predominantly Hispanic schools enforce a strict
dress code, compared to about one-third of predominantly White schools. In
View GAO-23-105348. For more information,
contact Jacqueline M. Nowicki at (617) 788-
0580 or nowickij@gao.gov.
Why GAO Did This Study
In recent years, researchers,
advocates, parents, and students have
raised concerns about equity in school
dress codes. Concerns have included
the detrimental effects of removing
students from the classroom for dress
code violations.
A committee report accompanying
H.R. 7614 included a provision for
GAO to study dress code discipline.
This report also addresses a request to
study informal removals. This report
examines (1) the characteristics of K-
12 dress codes across school districts
nationwide, and how Education
supports the design of equitable and
safe dress codes; (2) the enforcement
of dress codes, and how Education
supports equitable dress code
enforcement.
To examine characteristics of dress
codes, GAO analyzed a nationally
representative sample of public school
district dress codes. To assess the
enforcement of dress codes and how
Education supports school districts,
GAO analyzed Education data;
reviewed relevant studies on dress
code discipline; and interviewed
academic researchers and officials
from national organizations, school
districts, and Education.
What GAO Recommends
GAO is making four recommendations,
including that Education provide
resources to help districts design
equitable dress codes and collect and
disseminate information on the
prevalence and effects of informal
removals and non-exclusionary
discipline. Education described steps
to implement all four
recommendations.
addition, schools that enforce strict dress codes are associated with statistically
significant higher rates of discipline that removes students from the classroom
(e.g., suspensions). Further, an estimated 44 percent of dress codes outlined
“informal” removal policies, such as removing a student from class without
documenting it as a suspension. Education has recently noted challenges related
to informal removals in guidance documents but has no information on the
prevalence or impact of this emerging issue. Without information on the full range
of ways children are disciplined—including informal removals and non-
exclusionary discipline—Education’s efforts to provide resources on the equitable
enforcement of discipline will have critical gaps.
Page i GAO-23-105348 K-12 Education
Contents
GAO Highlights ii
Why GAO Did This Study ii
What GAO Recommends ii
What GAO Found ii
Letter 1
Background 4
Dress Codes Often Restrict Girls’ Clothing and Students’ Hair and
Head Coverings, and Education Does Not Have Resources on
Designing Equitable and Safe Dress Codes 10
Schools That Enforce Strict Dress Codes Suspend and Expel
More Students, and May Also Informally Remove Students for
Dress Code Violations 23
Conclusions 39
Recommendations for Executive Action 40
Agency Comments and Our Evaluation 41
Appendix I: Objectives, Scope, and Methodology 44
Appendix II: Technical Appendix for Regression Analyses 53
Appendix III: Snapshot of Selected School Districts That Revised Dress Codes 60
Appendix IV: Comments from the U.S. Department of Education 61
Accessible Text for Appendix IV: Comments from the U.S. Department of Education 64
Appendix V: GAO Contact and Staff Acknowledgments 67
Tables
Table 1: Population, Sample, and Completed Review Counts for
Our In Scope School Districts 46
Table 2: Created Variables Used in the Regression Analysis of
Department of Education’s School Survey on Crime and
Safety and Civil Rights Data Collection, School Years
2015-2016, and 2017-2018 57
Table 3: Variables Included in Our Regression Models Using the
Department of Education’s School Survey on Crime and
Safety and Civil Rights Data Collection, School Years
2015-2016, and 2017-2018 58
Page ii GAO-23-105348 K-12 Education
Table 4: Variables Included in Our Regression Model Using the
Department of Education’s School Survey on Crime and
Safety and Civil Rights Data Collection, 2017-2018 59
Table 5: Snapshot of Selected School Districts That Revised
Dress Codes 60
Figures
Figure 1: Examples of Items Prohibited by School Dress Code
Policies 5
Accessible Data for Figure 1: Examples of Items Prohibited by
School Dress Code Policies 5
Figure 2: Estimated Percentage of Schools Nationwide That
Report Enforcing a Strict Dress Code or Requiring
Uniforms, 1999-2020 8
Accessible Data for Figure 2: Estimated Percentage of Schools
Nationwide That Report Enforcing a Strict Dress Code or
Requiring Uniforms, 1999-2020 8
Figure 3: Estimated Percentage of Districts that Prohibit Clothing
Items Typically Worn by Girls or Boys 12
Accessible Data for Figure 3: Estimated Percentage of Districts
that Prohibit Clothing Items Typically Worn by Girls or
Boys 12
Figure 4: Estimated Percentage of Districts Prohibiting the
Exposure of Specific Body Parts 13
Accessible Data for Figure 4: Estimated Percentage of Districts
Prohibiting the Exposure of Specific Body Parts 13
Figure 5: Dress Code Excerpts: Examples of Rules Containing
Measurements 14
Accessible Data for Figure 5: Dress Code Excerpts: Examples of
Rules Containing Measurements 14
Figure 6: Dress Code Excerpts: Examples of Rules with
Subjective Language 15
Accessible Data for Figure 6: Dress Code Excerpts: Examples of
Rules with Subjective Language 15
Figure 7: Examples of religious and culturally significant head
coverings 16
Accessible Text for Figure 7: Examples of religious and culturally
significant head coverings 16
Figure 8: Culturally significant hairstyles and hair coverings 17
Accessible Data for Figure 8: Culturally significant hairstyles and
hair coverings 18
Page iii GAO-23-105348 K-12 Education
Figure 9: Estimated Percentage of School Districts Citing a
Particular Purpose for Their Dress Code, among the
Estimated 92 Percent of Districts That Stated a Purpose 19
Accessible Data for Figure 9: Estimated Percentage of School
Districts Citing a Particular Purpose for Their Dress
Code, among the Estimated 92 Percent of Districts That
Stated a Purpose 19
Figure 10: Estimated Percentage of Schools in Each Racial/Ethnic
Category That Report Enforcing a Strict Dress Code,
School Year 2017-18 25
Accessible Data for Figure 10: Estimated Percentage of Schools
in Each Racial/Ethnic Category That Report Enforcing a
Strict Dress Code, School Year 2017-18 25
Figure 11: Estimated Percentage of Public Schools That Report
Enforcing a Strict Dress Code, by Census Division,
School Year 2017-18 26
Accessible Data for Figure 11: Estimated Percentage of Public
Schools That Report Enforcing a Strict Dress Code, by
Census Division, School Year 2017-18 26
Figure 12: Estimated Percentage of Students Experiencing
Exclusionary Discipline in Schools That Enforce Strict
Dress Codes, School Year 2017-18 30
Accessible Data for Figure 12: Estimated Percentage of Students
Experiencing Exclusionary Discipline in Schools That
Enforce Strict Dress Codes, School Year 2017-18 30
Figure 13: Dress Code Excerpts: Examples of Removing Students
from Class 35
Accessible Data for Figure 13: Dress Code Excerpts: Examples of
Removing Students from Class 35
Figure 14: Non-Exclusionary Discipline Cited in Dress Codes
Violations, by Estimated Percent of Districts 37
Accessible Data for Figure 14: Non-Exclusionary Discipline Cited
in Dress Codes Violations, by Estimated Percent of
Districts 38
Abbreviations
APA American Psychological Association
CCD Common Core of Data
CRDC Civil Rights Data Collection
LEA local educational agency
NCES National Center for Education Statistics
Page iv GAO-23-105348 K-12 Education
NCSSLE National Center on Safe and Supportive Learning
Environments
OCR Office for Civil Rights
OSSS Office of Safe and Supportive Schools
REL Regional Educational Laboratory
SSOCS School Survey on Crime and Safety
This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.
Page 1 GAO-23-105348 K-12 Education
441 G St. N.W.
Washington, DC 20548
Letter
October 25, 2022
Congressional Addressees
Nearly every public school district in the nation requires students to
adhere to a dress code. School dress codes provide overall guidelines for
how students are expected to dress for school; establish rules about
clothes, hair, and accessories; and lay out disciplinary consequences for
violating the dress code. For example, students who violate a school’s
dress code may be asked to change clothes, be sent home, or be
suspended from school.
In recent years, researchers, advocates, parents, and students have
raised concerns that dress codes disproportionately focus on girls’
clothing and bodies and that exclusionary discipline—the practice of
removing students from the classroom—for dress code violations may
disproportionately harm Black and Hispanic students, among other
students.
1
Recent reports and research studies have garnered national
attention and have shed light on concerns with certain elements of dress
codes, including those that are unclear or overly strict, require expensive
purchases, or prohibit items associated with cultural or racial identity,
such as banning head coverings or traditionally Black hairstyles.
2
In
addition, some of these studies have noted that having different dress
codes for girls and boys can present obstacles for transgender and
nonbinary students. Some school districts have responded to dress code
controversies by revising their dress codes—sometimes citing a
commitment to equity and inclusion when doing so—or by switching to
uniforms.
A committee report accompanying the House bill for the Departments of
Labor, Health and Human Services, and Education, and Related
1
The federal data sources we cite in this report use the term “Hispanic or Latino” in their
data collection, which refers to a person of Cuban, Mexican, Puerto Rican, South or
Central American, or other Spanish culture or origin regardless of race. We use the term
Hispanic for the purposes of our reporting objectives.
2
National Women’s Law Center, Dress Coded: Black girls, bodies, and bias in D.C.
schools (2018), available at https://nwlc.org/resources/dresscoded; Dignity in Schools, A
Model Code on Education and Dignity: Presenting a human rights framework for schools
(2019), available at https://dignityinschools.org/modelcoded.
Letter
Page 2 GAO-23-105348 K-12 Education
Agencies Appropriations Act, 2021, included a provision for GAO to study
how school dress code and discipline policies are formulated and
executed across the country, ways in which dress codes may infringe
upon students’ civil rights, and promising practices related to dress code
discipline policies.
3
GAO was also asked to look at the topic of informal
removals by House Committee on Education and Labor Chairman Bobby
Scott and Representative A. Donald McEachin.
This report examines (1) the characteristics of K-12 dress codes across
school districts nationwide, and how the Department of Education
supports the design of equitable and safe dress codes; and (2) the
enforcement of dress codes, and how Education supports equitable dress
code enforcement.
To estimate the prevalence of dress code characteristics nationwide, we
analyzed publicly available dress code information from a nationally
generalizable, stratified random sample of school districts.
4
Using
Education’s Common Core of Data for school year 2020-21, we selected
a nationally representative sample of 236 public school districts and
systematically reviewed each district’s dress code using a structured data
collection instrument. We used a sampling strategy that accounted for
district demographics, size, and other variables to ensure certain
subpopulations of students were appropriately represented.
To obtain information on the enforcement of dress code discipline, we
analyzed data on dress codes and uniforms from Education’s School
Survey on Crime and Safety (school survey) for school years 2015-16
and 2017-18, the most recent available at the time of our review. We
matched the school survey with the 2015-16 and 2017-18 Civil Rights
Data Collection and Common Core of Data, and conducted generalized
linear regressions to explore associations between school-level
characteristics and policies. Such associations included enforcing a
“strict” dress code and rates of incidents of exclusionary discipline, such
as the percentage of students suspended, while controlling for other
factors such as school type and student demographics. Similarly, using
the 2017-18 data, we conducted generalized linear regressions to explore
associations between school-level characteristics and whether a school
3
H.R. Rep. No. 116-450, at 288 (2020).
4
Specifically, we stratified the sample into mutually exclusive strata that accounted for
districts’ number of students enrolled, urban classification (urban, suburban, or town/rural),
charter/non-charter status, and racial demographics. See appendix I for additional details.
Letter
Page 3 GAO-23-105348 K-12 Education
enforces a strict dress code, while controlling for other factors. We
conducted electronic data testing and obtained information from data
officials at Education, among other steps, to determine that these
datasets were reliable for these purposes.
Our models did not allow us to address causality, so we conducted a
targeted literature review to provide context for our findings. We identified
and reviewed relevant studies on discipline resulting from dress code
violations in K-12 public schools for the last 10 years (August 2011-
August 2021) that met our criteria.
5
We interviewed officials from three school districts that recently revised
their dress codes. We selected these districts for varying size, geographic
location, student demographics, and strategies used to revise their dress
codes (e.g., how data were collected, whether stakeholders were
consulted). Our interviews with school district officials are not
generalizable to all districts nationwide, but provide illustrative examples
of strategies districts may use when designing and revising dress code
policies. We also interviewed researchers and officials from national
organizations that conduct work on dress code discipline. Using social
media and outreach with a national organization, we also invited families
to participate in a brief questionnaire to obtain anecdotal perspectives
about their children’s dress codes/uniforms.
Finally, to obtain information on Education’s resources related to dress
code discipline, we reviewed relevant federal laws, regulations, guidance,
and documents, and interviewed agency officials. We compared
Education’s efforts with the agency’s strategic goals and objectives.
We conducted this performance audit from August 2021 to October 2022
in accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
5
To be included in our review, studies must use empirical student level data, relate to
student outcomes associated with dress code disciplinary actions, use rigorous statistical
methods, and be documented in a peer-reviewed journal.
Letter
Page 4 GAO-23-105348 K-12 Education
Background
According to the Education Commission of the States, as of December
2021, 24 states and the District of Columbia explicitly grant local districts
the power to establish dress codes.
6
Some states and localities have also
enacted rules about the content of K-12 dress codes, such as those that
prohibit race-based hair discrimination in educational settings. For
example, California’s Educational Code prohibits, among other things,
discrimination on the basis of race and ethnicity, and defines race
“inclusive of traits historically associated with race, including, but not
limited to, hair texture and protective hairstyles.”
Dress codes may be embedded in larger discipline documents or student
codes of conduct. They may prohibit specific articles of clothing,
accessories, hair styles, or makeup. In addition, dress codes may contain
subjective language about clothing, such as that clothing be “appropriate,”
or not be “excessively tight,” “distracting,” “revealing,” or “sexually
suggestive.” Dress code policies also sometimes differ based on
students’ sex or gender (see fig. 1).
6
We did not conduct a comprehensive review or analysis of state laws or policies as part
of this work.
Letter
Page 5 GAO-23-105348 K-12 Education
Figure 1: Examples of Items Prohibited by School Dress Code Policies
Accessible Data for Figure 1: Examples of Items Prohibited by School Dress Code Policies
Example of K-12 School Dress Code: Prohibited items may not be worn on campus or at school-sponsored
activities
Boys: Hats or durags; Bandanas; Muscle shirts; Inappropriate images or messages including: Foul language,
Anything that is sexually suggestive, or Promotion of drug or alcohol use
Girls: Visible piercings (other than ears); Visible bra strap(s); Thin (spaghetti) straps; Exposed cleavage;
Anything too tight; Exposed midriff; Skirt or shorts shorter than fingertips with arms fully extended; Leggings
worn alone as pants; High heel shoes
Both: Non-natural hair colors; Visible underwear; Holes in pants above knee; Open-toe shoes
Source: GAO summary of publicly-available dress codes; stock.adobe.com (base artwork). | GAO-23-105348
There are a variety of consequences for dress code violations. The
enforcement of dress codes can include requiring students to change
Letter
Page 6 GAO-23-105348 K-12 Education
clothes, in-school suspension, out-of-school suspension, and even
expulsion. Some examples of dress code enforcement have drawn
attention from the media. For a variety of examples of recent media
reports on dress code enforcement, see the text box below.
Letter
Page 7 GAO-23-105348 K-12 Education
Examples of dress code enforcement reported in the media from
April 2018 to June 2022
· A high school girl was told to “move around” for the school dean to
determine if her nipples were visible through her shirt. The student
was then instructed to put band aids on her chest.
· School staff drew on a Black boy’s head in permanent marker to
cover shaved designs in his hair.
· A female transgender student was told not to return to school until
she was following the school’s dress code guidelines for males.
· A high school girl was suspended for 10 days and prohibited from
attending her graduation ceremony for wearing a top that showed
her shoulders and back.
· Middle school girls were gathered at an assembly on dress code
and told they should not report inappropriate touching if they were
not following the dress code.
· A Black student was told he needed to remove his hair covering
(also called a durag) because an administrator said it was gang-
related.
· Two Asian American and Pacific Islander students were banned
from wearing leis and tupenus (cloth skirts)—cultural symbols of
celebration and pride—to their high school graduation.
Source: GAO review of selected news reports. | GAO 23 105348
RoleoftheDepartmentofEducationinSchoolDress
CodeEnforcement
Education collects school-reported data on whether public schools
enforce a strict dress code or require uniforms through its School Survey
on Crime and Safety (school survey). The school survey is designed to
provide estimates of school crime, discipline, disorder, programs, and
policies, including dress code and uniform policies. In its 2021 Digest of
Education Statistics, Education estimates that, for school year 2019-20,
nearly half of schools nationwide reported they enforce a strict dress code
and nearly one in five require uniforms. These practices have remained
relatively stable over time (see fig. 2).
Letter
Page 8 GAO-23-105348 K-12 Education
Figure 2: Estimated Percentage of Schools Nationwide That Report Enforcing a Strict Dress Code or Requiring Uniforms,
1999-2020
Accessible Data for Figure 2: Estimated Percentage of Schools Nationwide That Report Enforcing a Strict Dress Code or
Requiring Uniforms, 1999-2020
Academic year
Enforce a strict
dress code
(Lower bound)
Enforce a strict
dress code
(Estimate)
Enforce a strict
dress code
(Upper bound)
Require
uniforms
(Lower bound)
Require uniforms
(Estimate)
Require
uniforms
(Upper bound)
1999-2000
45.9
47.4
48.9
11
11.8
12.6
2003-04
53.9
55.1
56.3
13
13.8
14.7
2005-06
54.1
55.3
56.5
13
13.8
14.6
2007-08
53.6
54.8
56
16.8
17.5
18.2
2009-10
55.3
56.9
58.5
17.9
18.9
19.9
2013-14
56.9
58.5
60.1
19.1
20.4
21.7
2015-16
51.9
53.1
54.3
20.1
21.5
22.9
2017-18
47.5
48.8
50.1
18.9
19.8
20.7
2019-20
42.38
43.7
45.02
17.66
18.8
19.94
Source: National Center for Education Statistics, Digest of Education Statistics, October 2021. | GAO-23-105348
Education’s Office of Elementary and Secondary Education has an Office
of Safe and Supportive Schools (OSSS) that addresses the health and
well-being of students and school safety, security, and emergency
management and preparedness. OSSS administers, coordinates, and
recommends policy in addition to managing grant programs and technical
Letter
Page 9 GAO-23-105348 K-12 Education
assistance centers that address the overall safety and health of school
communities. One of its technical assistance centers, the National Center
on Safe and Supportive Learning Environments (NCSSLE) focuses on
improving student supports and academic enrichment by providing
technical assistance and support to states, districts, schools, and the
public on school climate and related topics. NCSSLE has developed
resources on school discipline and creating positive school climates, and
also offers related resources from external parties, including some
resources related to dress code discipline.
7
Education’s Office for Civil Rights (OCR) is responsible for enforcing
certain federal civil rights laws that prohibit discrimination in schools and
other programs or activities that receive federal assistance from
Education, such as
· Title VI of the Civil Rights Act of 1964 (Title VI), which prohibits
discrimination on the basis of race, color, or national origin by
recipients of federal funding;
8
· Title IX of the Education Amendments of 1972 (Title IX), which
prohibits sex discrimination in education programs that receive federal
funding;
9
and
· Section 504 of the Rehabilitation Act of 1973 (Section 504) which
prohibits discrimination on the basis of disability by recipients of
federal funding.
10
Additionally, OCR has responsibilities under Title II of the Americans with
Disabilities Act of 1990 (Title II), which prohibits discrimination on the
basis of disability by public entities (such as public school districts, public
7
Information about NCSSLE’s resources are available on its website:
https://safesupportivelearning.ed.gov/.
8
42 U.S.C. §§ 2000d – 2000d-7. Although Title VI does not prohibit discrimination on the
basis of religion, according to Education, Title VI protects students of any religion from
discrimination, including harassment, based on a student’s actual or perceived shared
ancestry or ethnic characteristics, or citizenship or residency in a country with a dominant
religion or distinct religious identity.
9
20 U.S.C. §§ 1681 – 1689
10
29 U.S.C. § 794.
Letter
Page 10 GAO-23-105348 K-12 Education
colleges and universities, and public libraries), whether or not they
receive federal financial assistance.
11
DressCodesOftenRestrictGirls’Clothingand
Students’HairandHeadCoverings,and
EducationDoesNotHaveResourceson
DesigningEquitableandSafeDressCodes
DressCodesOftenRestrictGirls’ClothingandInclude
RulesThatNecessitateMeasuringStudents’Bodiesand
Clothes
Rules related to girls’ clothing and bodies. Nearly all K-12 public
school districts (an estimated 93 percent) have a policy on student dress,
according to our nationally generalizable review of district policies.
12
These dress codes more frequently restrict items typically worn by girls—
such as short skirts, spaghetti strap tank tops, and leggings—than those
typically worn by boys—such as muscle shirts (see fig. 3). An estimated
90 percent of dress codes prohibit clothing items typically associated with
girls compared to 69 percent that prohibit items typically associated with
boys.
13
Dress codes we reviewed include statements such as “halter or
strapless tops, and skirts or shorts shorter than mid-thigh are also
prohibited” and “yoga pants or any type of skin tight attire may not be
worn by itself. It may be worn underneath [other clothing] that is long
enough to ensure modesty.” Some parents who responded to our online
questionnaire expressed appreciation for aspects of their children’s dress
11
42 U.S.C. §§ 12131 – 12134.
12
Unless otherwise noted, all estimates from our review of publicly available district dress
codes have a margin of error of plus or minus 7 percentage points or less, at the 95
percent confidence level. The percentage estimates of school districts are based on
district information that was publicly available on school district websites from April-May
2022.
13
We identified clothing that is typically associated with girls or boys through a review of
prior research related to school dress codes, gendered language in our sample of dress
codes, and a review of “girls” and “boys” sections of the websites of national children’s
clothing retailers.
Letter
Page 11 GAO-23-105348 K-12 Education
Q&A: What do families think about school
dress codes and uniforms?
We asked families to respond to a
questionnaire about the dress code and
uniform policies in their children’s school.
Below are selected responses that represent
a range of views expressed:
· “My girls definitely feel anger towards the
school for not educating the boys and
making [the girls] aware every day what
they wear can be a distraction to the
boys.”
· “They love the dress code/uniform policy.
It is one less thing they have to think
about when they wake up in the morning
and they don’t have to worry about not
fitting in.”
· “They don’t like the ‘boys wear pants and
girls wear skirts’ type [of] language used
in the dress code. They want to be able
to wear anything they choose.”
· “It’s helped reinforce our own family
opinions that our 12 year old doesn’t
need to go to school with her stomach
showing.”
Source: GAO questionnaire. | GAO-23-105348
codes; others had concerns about dress code rules focused on clothing
typically worn by girls (see sidebar).
14
14
Using social media and outreach with a national organization, we invited families to
participate in a brief questionnaire to obtain anecdotal perspectives about their children’s
dress codes/uniforms. We received 47 responses in March and April 2022.
Letter
Page 12 GAO-23-105348 K-12 Education
Figure 3: Estimated Percentage of Districts that Prohibit Clothing Items Typically Worn by Girls or Boys
Accessible Data for Figure 3: Estimated Percentage of Districts that Prohibit Clothing Items Typically Worn by Girls or Boys
Percentage of dress codes prohibiting
Category
Lower bound
Estimate
Upper bound
At least one clothing item typically worn by boys
62.2
68.8
75.5
Muscle shirts
39.1
46.2
53.4
Sagging pants
48.7
55.6
62.5
At least one clothing item typically worn by girls
a
85.3
90.3
94
Midriff-bearing tops and/or spaghetti straps
70.3
75.9
81.6
Short skirts or shorts
69.3
75.4
81.6
Leggings worn as pants
24.4
30.7
37
Source: GAO review of school dress codes; stock.adobe.com (base artwork). | GAO-23-105348
a
Prohibited items typically worn by girls also include the following items not shown: tops with low cut
necklines, shoes with heels of a certain height, nylon and spandex, and clothing that is sheer or
transparent/translucent.
Most dress codes stipulate that students’ clothing must cover specific
body parts; these restrictions more frequently apply to clothing typically
Letter
Page 13 GAO-23-105348 K-12 Education
worn by girls, such as halter or crop tops. For example, most dress codes
(67 percent) prohibit clothing that exposes a student’s midriff. We also
estimate about a quarter of district dress codes specifically prohibit the
exposure of “cleavage,” “breasts,” or “nipples” (see fig. 4).
Figure 4: Estimated Percentage of Districts Prohibiting the Exposure of Specific
Body Parts
Accessible Data for Figure 4: Estimated Percentage of Districts Prohibiting the
Exposure of Specific Body Parts
Percentage of dress codes
Category
Lower
bound
Estimate
Upper
bound
Midriff, stomach, or abdomen
60.5
66.9
73.3
Cleaveage, breasts, or nipples
17.5
23.7
29.8
Buttocks
6.7
10.6
15.5
Shoulder
4.4
7.3
11.4
Armpits
2.3
4.7
8.6
Thighs
1
2.8
6.1
Knees
0
0
1.4
Source: GAO review of district dress codes. | GAO-23-105348
Rules requiring measurements and subjective interpretation. An
estimated 60 percent of districts use measurements to determine if
student clothing is permitted, based on our generalizable sample of
districts (see fig. 5).
Letter
Page 14 GAO-23-105348 K-12 Education
Figure 5: Dress Code Excerpts: Examples of Rules Containing Measurements
Accessible Data for Figure 5: Dress Code Excerpts: Examples of Rules Containing Measurements
· Rural district: Shorts should have a minimum 3 1/2 inch inseam
· Suburban charter district: No straps smaller than three adult fingers
· Suburban district: Non-button shirts must not exceed two inches from the bottom of neckline. Torn
areas must be less than the size of a tennis ball... If pants have tears 2 inches above the knee, student
must wear opaque leggings underneath. Dresses, shorts and skirts must be no shorter than two (2)
inches above the knee at all times
Source: GAO review of dress codes. | GAO-23-105348
Officials we spoke with from national organizations raised concerns that
measurement provisions in dress codes may lead to adults touching
students’ bodies to measure clothes. Officials at national organizations
and district officials noted that having staff determine if students’ shorts or
skirts met the required length was embarrassing to students, particularly if
this was done in front of their peers. In our review, we found examples of
dress codes that required students to move or stand in a specific way for
staff to check if their clothing conforms to the stated measurement rule.
For example, one dress code stated, “The test: No bare midsection or
back is revealed when arms are stretched over head.” Officials from
national organizations also raised concerns that aspects of dress codes
may have implications for student privacy. For example, we found
examples of dress codes that require students to wear undergarments,
such as “for females: bras must be worn.”
Moreover, almost all district dress code policies (an estimated 93 percent)
contain rules with subjective language that leave decisions about dress
code compliance open to interpretation. Commonly used subjective
phrases, such as “revealing” or “immodest” clothing, often apply to
standards of appearance typically associated with girls and women (see
fig. 6).
Letter
Page 15 GAO-23-105348 K-12 Education
Figure 6: Dress Code Excerpts: Examples of Rules with Subjective Language
Accessible Data for Figure 6: Dress Code Excerpts: Examples of Rules with Subjective Language
· Urban district: Unkempt appearances must be corrected before returning to class. Extremes of any kind
of fad or wearing apparel which call undue attention to the person are not acceptable… Modesty is the
expectation for dress on campus and at school-sponsored events.
· Suburban charter district: Whenever any element of physical appearance or grooming, even if it is
allowable under the school’s current rules becomes a distraction... it may be considered outside of
dress code.
· Rural district: Students will observe modesty, appropriateness, and neatness in clothing and personal
appearance.
· Rural district: Clothing worn in a very sloppy manner or that is in the extreme cannot be tolerated in a
situation where learning is the primary objective.
Source: GAO review of dress
According to researchers and officials at national organizations, rules that
are open to interpretation may also be disproportionately applied to
vulnerable student groups including LGBTQI+ students, Black students,
and students with disabilities.
15
About half of dress codes (an estimated
46 percent) have rules about the way clothing fits students, prohibiting
clothing that is “too tight” or “too loose.” Researchers have raised
concerns that this type of language may be enforced unequally based on
students’ body type or body maturity and risks causing embarrassment
and body shame for students.
15
While a number of variations on this acronym are currently in use to describe individuals
with diverse sexual orientations and gender identities, in this report, we define LGBTQI+
as lesbian, gay, bisexual, transgender, queer, questioning, or intersex. The “plus” is meant
to be inclusive of identities that may not be covered by the acronym LGBTQI+, including
asexual, nonbinary, and individuals who identify their sexual orientation or gender identity
in other ways.
Letter
Page 16 GAO-23-105348 K-12 Education
MostDistrictsHaveRulesaboutHeadCoveringsand
Students’Hair,andFewSpecifyReligious,Cultural,or
MedicalExemptions
Our generalizable analysis of dress codes found that over 80 percent of
districts prohibit head coverings such as hats, hoodies, bandanas, and
scarves; only one-third of these dress codes specify that they allow
religious exemptions; and few specify cultural or disability/medical
exemptions.
16
However, some head coverings have religious or cultural
significance for students (see fig. 7).
Figure 7: Examples of religious and culturally significant head coverings
Accessible Text for Figure 7: Examples of religious and culturally significant head coverings
Students may wear head coverings as a form of religious observance or a way to express cultural identity.
Here are three examples:
· picture of two women wearing a hijab
· picture of children wearing a kippah
· picture of child wearing a kufi cap
Source: GAO summary of interviews with national organizations; stock.adobe.com (photos). | GAO-23-105348
16
An estimated 5 percent of district dress codes specified exemptions for cultural items.
An estimated 30 percent of dress codes specify exemptions for medical or disability-
related items, such as allowing students with certain medical conditions to wear a baseball
cap. We reviewed publicly available dress code information from school districts; we did
not ask school district officials about religious, cultural, or disability-related exemptions.
The fact that a dress code did not specify an exemption does not necessarily mean that
the district does not allow for an exemption.
Letter
Page 17 GAO-23-105348 K-12 Education
In addition, most dress codes (an estimated 59 percent) contain rules
about students’ hair, hairstyles, and hair coverings, and these rules may
disproportionately impact Black students, according to researchers and
district officials we interviewed. For example, many districts (an estimated
44 percent) ban hair wraps, with some specifically naming durags or other
styles of hair wraps. In addition, one in five dress codes (an estimated 21
percent) include rules on student hair with subjective language such as,
“hair must look natural, clean, and well-groomed” or say students’ hair
must not be “distracting” or “extreme.”
We found a small number of dress codes with restrictions on hair length
(an estimated 2 percent) or dress codes that prohibited shaved lines in
hair (4 percent). Finally, we found examples of dress codes with rules
specific to natural, textured hair, which researchers have noted
disproportionately affect Black students. For example, one district
prohibited hair with “excessive curls” and another stated that “hair may be
no deeper than two inches when measured from the scalp.” (see fig. 8).
Figure 8: Culturally significant hairstyles and hair coverings
Education Case: Investigation of Alleged
Racial Discrimination in an Arizona School
In 2017, Education’s Office for Civil Rights
(OCR) investigated whether a student was
discriminated against on the basis of race
when it was reported that he was repeatedly
told by a teacher, then later the school
director, to cut or change his hair after
wearing an afro. According to OCR
documents, the school director said the afro
violated the school’s policy against “trendy
hairstyles” and school officials also cited
reasons not written in an official policy.
OCR found that the reasons given were
pretexts for discrimination, as the school
allowed “long hair that grows down (the
natural hair growth direction for most White
people), but not long hair that grows out/up
(the natural hair growth direction for most
African American people).” As a result, the
school voluntarily agreed to enter into a
resolution agreement to resolve the matter.
Source: GAO review of Department of Education documents.
| GAO-23-105348
Letter
Page 18 GAO-23-105348 K-12 Education
Accessible Data for Figure 8: Culturally significant hairstyles and hair coverings
Hairstyles, such as cornrows, locks, twists, afros, bantu knots, are an
expression of Black identity, culture, religion, and history.
In addition, the elasticity and texture of Black hair can make it more
susceptible to breaking. Black students may want to wear “protective
hairstyles” like braids and twists to maintain healthy hair. These
hairstyles, and hair coverings such as a durag, can be worn without
manipulating or damaging hair.
Source: GAO summary of NAACP fact sheet; stock.adobe.com (photos). | GAO 23-105348
SchoolDressCodesAreCommonlyAimedatPromoting
SafetyofStudents
A commonly stated purpose of school dress codes is student safety and
security, with an estimated 73 percent of dress codes citing this as a goal
for the design of their dress codes (see fig. 9). Education officials noted
that schools can use dress codes to address safety and security, and
district officials stated that aspects of their dress codes were aimed at
promoting student safety. For example, officials in two of the districts we
interviewed said that in their districts, prohibiting hats and head coverings
is a safety measure to ensure administrators can readily identify students.
Letter
Page 19 GAO-23-105348 K-12 Education
Figure 9: Estimated Percentage of School Districts Citing a Particular Purpose for Their Dress Code, among the Estimated 92
Percent of Districts That Stated a Purpose
Accessible Data for Figure 9: Estimated Percentage of School Districts Citing a Particular Purpose for Their Dress Code,
among the Estimated 92 Percent of Districts That Stated a Purpose
· Suburban district: The student dress code is intended to create and preserve a positive climate for
teaching and learning… and preserve school order and safety.
· Urban charter district: As a security measure, uniforms provide us the ability to identify all of our
scholars inside or outside the building
· Suburban charter district: Appropriate dress is important in providing a safe, distraction-free
environment.
Frequency of each stated purpose in district dress codes
(Dress codes often contained more than one reason in the purpose statements)
Letter
Page 20 GAO-23-105348 K-12 Education
Frequently stated reasons
Lower bound
Estimate
Upper bound
Eliminating distractions
66.8
73
79.3
Student safety
66.3
72.8
79.2
Enhancing student academics or professionalism
52.3
59.6
66.8
Improving student behavior
12.5
17.8
24.3
Allowing student expression
12.3
17.4
23.6
Describing required uniforms
13.2
17.2
21.9
Prohibiting offensive choices
11.1
16.4
23
De-emphasizing socioeconomic disparities
4.3
7.1
11
Preventing gang activity
2.9
5.3
8.7
Source: GAO review of district dress codes. | GAO-23-105348
However, a range of stakeholders we interviewed—from school districts,
national organizations, and researchers—noted that aspects of dress
codes may inadvertently contribute to a less safe and secure environment
for students, particularly girls, for several reasons:
· The focus on clothing typically associated with girls reinforces the
harmful view that girls are responsible for distracting boys.
17
Researchers have pointed out that this unfairly burdens girls.
· Dress codes that focus on girls’ clothing and bodies contribute to
shifting the burden of being harassed from the perpetrator to the
victim, thus potentially creating an environment that is less safe for
girls.
18
· Dress codes that involve measurements or that may necessitate
adults touching students, leave students, often girls, more vulnerable
to inappropriate touching and sexual harassment.
17
In its 2021 primer on “Students Experiencing Inattention and Distractibility,” the
American Psychological Association (APA) notes that educators should not make
assumptions or claims about what is causing students’ inattention. On the topic of
sexualization of girls, the APA states that “parents can teach boys to value girls as friends,
sisters and girlfriends, rather than as sexual objects.”
18
The APA reports that sexualizing clothing may be a factor in harassment of girls. The
APA report underscores that girls do not “cause” harassment or abusive behavior by
wearing “sexy” clothes; and that no matter what girls wear, they have the right to be free of
sexual harassment and boys and men can and should control their behavior. American
Psychological Association, Task Force on the Sexualization of Girls, Report of the APA
Task Force on the Sexualization of Girls (2007), retrieved from
http://www.apa.org/pi/women/programs/girls/report-full.pdf.
Letter
Page 21 GAO-23-105348 K-12 Education
EducationDoesNotProvideInformationonHowto
DesignEquitableandSafeDressCodes
Education does not make resources available on creating equitable dress
code policies; however, researchers and officials from national
stakeholder organizations and school district officials we spoke with said
such information would be helpful. For example, officials from national
organizations and researchers said that districts and schools could
benefit from examples of more equitable dress codes that have gender-
neutral and gender-inclusive language. Officials in our three selected
school districts—all of which recently revised their dress codes to
promote equity—noted challenges during the revision process (see
appendix III for more information on these selected districts).
19
Specifically, they noted that districts can have difficulty finding guidance
or best practices on designing equitable dress codes. They all said that a
key reason for revising their dress codes was that girls felt unfairly
burdened by the previous policies. In addition, they and officials from
national stakeholder organizations told us they are concerned about
language in dress codes that is not inclusive of all gender identities.
20
Research on dress code policies bear out these concerns. According to a
2019 national school climate survey, 18 percent of LGBTQI+ youth
reported that their school prevented them from wearing specific clothing
because it did not fit with the school’s perception of clothing appropriate
for their gender.
21
In our review of dress codes, we found that an
estimated 15 percent of districts’ dress codes specify different rules for
clothing, accessories, or hairstyles based on students’ sex, such as “no
fingernail polish or makeup is allowed on male students.” None of the
dress codes we reviewed with sex-based rules explicitly protect
19
One district aimed to increase socioeconomic equity by implementing uniforms; the
other two districts included changes such as eliminating gender-based language, limiting
subjective language, and increasing awareness of cultural identities.
20
Gender identity can be defined as a person’s innate, deeply felt psychological sense of
gender, which may or may not correspond to the person’s sex assigned at birth. Gender
identity is distinct and separate from sexual orientation.
21
Responding to the survey were 16,713 youth who identify as LGBTQI+ between the
ages of 13 and 21. The survey was administered between April and August 2019. J.G.
Kosciw, C.M. Clark, N.L. Truong, and A.D. Zongrone, The 2019 National School Climate
Survey: The experiences of lesbian, gay, bisexual, transgender, and queer youth in our
nation’s schools (New York: GLSEN, 2020).
Letter
Page 22 GAO-23-105348 K-12 Education
transgender or nonbinary students’ ability to dress according to their
gender identity.
Education has guidance on ways schools can support racial, cultural, and
gender equity, but this guidance does not explicitly address dress codes.
When we asked Education officials about providing districts or schools
with information to help them design equitable dress code policies, the
officials responded that the agency is committed to working on a range of
important topics and continues to consider a range of options to provide
helpful resources to stakeholders and the public. However, as of
September 2022, Education officials had not provided any additional
information on this topic. Resources to help districts and schools design
more equitable dress codes would align with Education’s goals to support
and build schools’ capacity to promote positive, inclusive, safe, and
supportive school climates in a nondiscriminatory manner.
Education also has a range of resources available on safety, security, and
school climate, but these publications contain limited information about
dress codes.
22
For example, Education’s National Center on Safe and
Supportive Learning Environments (NCSSLE) states that a positive
school climate “reflects attention to fostering social and physical safety,
providing support that enables students and staff to realize high
behavioral and academic standards as well as encouraging and
maintaining respectful, trusting, and caring relationships throughout the
school community.”
23
However, NCSSLE’s resources related to school
22
U.S. Department of Education Office for Civil Rights, Fact Sheet: Supporting Intersex
Students (October 2021),
https://www2.ed.gov/about/offices/list/ocr/docs/ocr-factsheet-intersex-202110.pdf; U.S.
Department of Education Office for Civil Rights and Department of Justice Civil Rights
Division, Fact Sheet: Confronting Discrimination Based on National Origin and
Immigration Status (August 2021),
https://www2.ed.gov/about/offices/list/ocr/docs/confronting-discrimination-national-origin-i
mmigration-status; U.S. Department of Education, Supporting Transgender Youth in
School (June 2021),
https://www2.ed.gov/about/offices/list/ocr/docs/ed-factsheet-transgender-202106.pdf.
23
U.S. Department of Education, Office of Safe and Healthy Students, Quick guide on
making school climate improvements (Washington, D.C.: 2018),
http://safesupportivelearning.ed.gov/SCIRP/Quick-Guide.
Letter
Page 23 GAO-23-105348 K-12 Education
safety or school climate improvement do not have information on the
design or implementation of dress codes.
24
Because many districts view dress codes as facets of larger safety,
security, and school climate policies, incorporating dress code information
and examples into existing safety and security resources could better
enable Education to efficiently provide information to districts and schools
that enhances social and physical safety for all students, a key agency
goal. Education officials stated that OCR and NCSSLE would work
together to support, among other things, the possible development of
additional discipline resources that Education would make available.
However, as of September 2022, Education had not provided dress code
information in agency resources on safe and secure schools.
By providing resources to districts and schools on the design of dress
codes that includes information on equity and safety, Education could
further support its mission to provide equal access to educational
opportunity. In the absence of this information, Education may miss an
opportunity to help districts create or revise dress codes to promote a
safe, supportive learning environment that embraces all students.
SchoolsThatEnforceStrictDressCodes
SuspendandExpelMoreStudents,andMay
24
Education officials noted that the NCSSLE has some resources available on its website
that were not developed by Education, and two of these resources mention dress codes.
For example, a 2014 report on school discipline provides dress code violation examples
and recommendations. See E. Morgan, N. Salomon, M. Plotkin, and R. Cohen, The
School Discipline Consensus Report: Strategies from the Field to Keep Students Engaged
in School and Out of the Juvenile Justice System (New York: The Council of State
Governments Justice Center, 2014). However, these resources do not represent
Education’s official positions.
Letter
Page 24 GAO-23-105348 K-12 Education
AlsoInformallyRemoveStudentsforDress
CodeViolations
SchoolswithLargerProportionsofBlackandHispanic
StudentsandSchoolsintheSouthAreMoreLikelyto
EnforceaStrictDressCode
According to our analysis of Education’s 2017-18 School Survey on
Crime and Safety (school survey) data, schools with higher percentages
of Black and Hispanic students are more likely to enforce strict dress
codes, holding other school characteristics constant.
25
This analysis
shows that the likelihood of a school enforcing a strict dress code
increases as the percentage of Black and Hispanic students increases.
As shown in figure 10, more than four in five predominantly Black schools
(schools where Black students comprise at least 75 percent of the student
body) and nearly two-thirds of predominantly Hispanic schools enforce a
strict dress code.
25
We performed a logistic regression analysis to hold other school characteristics—such
as geographic region, grade level, and school type—constant. Our analysis was limited by
the collinearity of race and poverty, as race and poverty are closely linked. In our analysis,
it was not possible to identify how much a school’s likelihood of enforcing a strict dress
code is attributable to race versus poverty.
Letter
Page 25 GAO-23-105348 K-12 Education
Figure 10: Estimated Percentage of Schools in Each Racial/Ethnic Category That
Report Enforcing a Strict Dress Code, School Year 2017-18
Accessible Data for Figure 10: Estimated Percentage of Schools in Each
Racial/Ethnic Category That Report Enforcing a Strict Dress Code, School Year
2017-18
Percentage of schools
Category
Lower
bound
Estimate
Upper
bound
Average for all U.S. schools
n/a
48.8
n/a
Predominantly Black schools
72.7
81
87.7
Predominantly Hispanic schools
61.6
62.8
64.0
Predominantly White schools
35.0
35.5
36.0
Source: GAO analysis of the Department of Education 2017-2018 School Survey on Crime and
Safety (SSOCS). | GAO-23-105348
Notes: Schools “predominantly” of a certain race/ethnicity are those where students of that particular
race/ethnicity make up 75 percent or more of the student population. We could not report on schools
predominantly enrolling Asian/Pacific Islander, Native American/Alaskan Native, or Multirace students
due to insufficient data. In addition, Hispanic students can be any race, but in Education’s data,
Hispanic is considered an ethnicity exclusive of race. Using a 95 percent confidence interval, the
margin of error for each school group is within +/- 8 percentage points.
A higher percentage of schools in the South enforce a strict dress code
with just over 70 percent of schools in the West South Central states
enforcing a strict dress code (see fig. 11). In contrast, less than 30
percent of schools in the West North Central states and in New England
enforced a strict dress code.
Q&A: What does “enforce a strict dress
code” mean?
In this report, schools that “enforce a strict
dress code” are identified and reported using
Education’s School Survey on Crime and
Safety, which asks school administrators,
“During the 2017-18 school year, was it a
practice of your school to enforce a strict
dress code?” The survey does not define “a
strict dress code” so school administrators
could interpret this question differently.
Enforcing a strict dress code and requiring
uniforms are addressed separately in the
survey. Regarding uniforms, the survey asks
“Was it a practice of your school to require
students to wear uniforms?” The data show
that, of schools that require a uniform, an
estimated 88 percent (95 percent confidence
interval 82.6-91.9) also report enforcing a
strict dress code. Of schools with strict dress
codes, only about 36 percent (95 percent
confidence interval 32.5-39.0) also report
requiring a uniform.
Source: GAO analysis of Education’s School Survey on
Crime and Safety. | GAO-23-105348
Letter
Page 26 GAO-23-105348 K-12 Education
Figure 11: Estimated Percentage of Public Schools That Report Enforcing a Strict Dress Code, by Census Division, School
Year 2017-18
Accessible Data for Figure 11: Estimated Percentage of Public Schools That Report Enforcing a Strict Dress Code, by Census
Division, School Year 2017-18
Census Division
Percentage of schools
enforcing a strict dress
code
Pacific (Alaska, Hawaii, Wash., Ore., Calif.)
44
Mountain (Mont., Idaho, Wyo., Nev., Utah, Colo., Ariz., New Mex.)
47
West North Central (N. Dak., S. Dak., Minn., Neb., Iowa., Kan., Mo.)
29
East North Central (Mich., Wisc., Ill., Ind., Ohio)
43
Mid Atlantic (N.Y., Penn., N.J.)
41
New England (Maine, N.H., Vt., Mass., Conn., R.I.)
28
West South Central (Texas, Okla., Ark., La.)
71
East South Central (Ken., Tenn., Miss., Ala.)
65
South Atlantic (Md., Del., W. Va., Va., N.C., S.C., Ga., Fla.)
57
Source: GAO analysis of the 2017-2018 School Survey on Crime and Safety (SSOCS) administered by the Department of Education; U.S. Census
Bureau (base map). | GAO-23-105348
Note: Using a 95 percent confidence interval, the margin of error for each Census division is within +/-
2 percentage points.
Similarly, when considering the four primary regions of the country (West,
Midwest, Northeast, and South), our regression analysis found that
Letter
Page 27 GAO-23-105348 K-12 Education
schools located in the South are estimated to be more than twice as likely
to enforce strict dress codes than schools in the Northeast.
26
In addition, national data show that schools that enforce strict dress
codes have other characteristics that differ from schools that do not
enforce strict dress codes. According to our regression analysis, schools
that are more likely to enforce a strict dress code have older students
(grades six and above). A higher percentage of large schools—schools
with more than 1,000 students—enforce a strict dress code.
27
An
estimated 71 percent of charter schools enforce strict dress codes
compared to 47 percent of non-charter schools. In addition, schools that
enforce strict dress codes also enroll higher percentages of students
eligible for free or reduced-price lunch—a proxy for students living in
poverty.
28
Officials from national organizations and districts we
interviewed noted that strict dress codes (and uniform requirements)
could pose challenges for low-income families who may struggle to buy
specific clothing items or afford certain hairstyles.
29
We analyzed schools
that require uniforms separately; see text box below.
26
Our odds ratio models estimate the likelihood that schools located in one region of the
country report enforcing a strict dress code, compared to schools located in the Northeast.
There are four Census regions and nine divisions. The West Census region is comprised
of the Pacific and Mountain divisions; the South region is comprised of the West South
Central, East South Central, and South Atlantic divisions; the Northeast region is
comprised of the Middle Atlantic and New England Regions; the Midwest region is
comprised of the West North Central and East North Central divisions. We estimated that
the odds of a school in the South having a strict dress code is 2.7 times the odds of a
school in the Northeast, and the 95 percent confidence interval of this estimate is 1.8-4.1.
27
An estimated 33 percent of small schools (1 to 200 students) report enforcing a strict
dress code compared to 50 percent of medium schools (201 to 1,000 students) and 56
percent of large schools (1,001 or more students). Using a 95 percent confidence interval,
the margin of errors for these estimates is within +/- 2 percentage points.
28
Using a 95 percent confidence interval, the margin of errors for these estimates on
charter schools is within +/- 2 percentage points.
29
Proponents of uniforms in public schools note that these policies can be a cost-savings
for families. However, officials we interviewed from national organizations stated that
uniforms can cause added expenses. For example, uniforms can be expensive and some
schools allow “dress down days or events” so families feel the need to purchase two sets
of clothing for students.
Letter
Page 28 GAO-23-105348 K-12 Education
Characteristics of Schools That Require Uniforms
A higher percentage of schools with the following characteristics require uniforms:
· More Black or Hispanic students: An estimated 72 percent of predominantly
Black and 52 percent of predominantly Hispanic schools require uniforms, as
compared to 2 percent of predominantly White schools. Schools “predominantly” of
a certain race/ethnicity are those where students of that particular race/ethnicity
make up 75 percent or more of the student population.
· Urban: An estimated 40 percent of schools in urban areas require students to wear
uniforms, as compared to 18 percent of schools in suburban areas and 7 percent in
rural areas.
· Elementary schools: An estimated 23 percent of elementary, 18 percent middle,
and 10 percent of high schools require students to wear uniforms.
· Charter schools: An estimated 64 percent of charter schools require students to
wear uniforms compared to 17 percent of non-charter schools.
Source: GAO analysis of the Department of Education’s School Survey on Crime and Safety for school year 2017-18; Rawpixel.com/stock.adobe.com. | GAO-23-105348
Note: Using a 95 percent confidence interval, the margin of errors for these estimates on uniforms is
within +/- 2 percentage points.
SchoolsThatEnforceStrictDressCodesAreAssociated
withHigherRatesofExclusionaryDiscipline,and
EducationDoesNotHaveGuidanceThatAddresses
DisparitiesinDisciplineEnforcement
Schools that enforce strict dress codes are associated with statistically
significant higher rates of exclusionary discipline—that is, practices that
remove students from the classroom, such as in-school suspensions, out-
of-school suspensions, and expulsions (see fig. 12).
30
This is true even
after controlling for student demographics, school type, size, geography,
and measures of school climate, such as levels of disorder and the
presence of security personnel (see appendix II for more information on
our regressions).
30
“Exclusionary discipline,” or any action that removes students from the learning
environment, includes more severe disciplinary measures such as suspensions and
expulsions, but it can also include sending students to the principal’s office or any other
action that takes a student out of the learning environment.
Exclusionary Discipline in Dress Codes
“Students who violate the [dress code] will not
be admitted to class and may be suspended
from school.”
Letter
Page 29 GAO-23-105348 K-12 Education
In our review of districts’ dress codes, we
found that an estimated 61 percent of dress
codes allow for the removal of students from
class for violations. Dress codes listed both
formal removal from the learning environment
(i.e., in-school and out-of-school suspensions)
and informal removal policies (i.e., students
sent home).
Source: GAO review of school district dress codes. |
GAO-23-105348
Letter
Page 30 GAO-23-105348 K-12 Education
Figure 12: Estimated Percentage of Students Experiencing Exclusionary Discipline in Schools That Enforce Strict Dress
Codes, School Year 2017-18
Accessible Data for Figure 12: Estimated Percentage of Students Experiencing Exclusionary Discipline in Schools That
Enforce Strict Dress Codes, School Year 2017-18
Percentage of students experiencing disciplinary action
Category
Schools that
do not enforce
a strict dress
code (lower
bound)
Schools that do
not enforce a
strict dress code
(estimate)
Schools that
do not enforce
a strict dress
code (upper
bound)
Schools that
enforce a strict
dress code
(lower bound)
Schools that
enforce a strict
dress code
(estimate)
Schools that
enforce a strict
dress code
(upper bound)
In-school
suspension
3.9
4.2
4.6
6
6.52
7
Out-of-school
suspension
3.5
3.7
3.9
5
5.4
5.9
Expulsions
0.1
0.1
0.15
0.18
0.2
0.26
Source: GAO analysis of School Survey on Crime and Safety (SSOCS) and Civil Rights Data Collection (CRDC) data for school year 2017-18. | GAO-
23-105348
Research shows exclusionary discipline is associated with short and long-
term negative outcomes for students, including increased risk for failing
standardized tests and increased rates of drop outs and incarceration.
31
31
One study showed that a single in-school suspension is predictive of significant risk for
academic failure (greater than 25 percent chance of failure) on a state-wide standardized
test, while controlling for individual and school level characteristics. See Danielle Smith,
Nickolaus A. Ortiz, Jamilia J. Blake, Miner Marchbanks III, Asha Unni, and Anthony A.
Peguero. “Tipping Point: Effect of the Number of In-school Suspensions on Academic
Failure,” California Contemporary School Psychology, 25 (May 2020): 466-475.
Letter
Page 31 GAO-23-105348 K-12 Education
For example, students who have been suspended are more likely to drop
out of school and become involved in the juvenile justice system than
their peers.
32
Studies, including our prior work, have shown disparities in who typically
gets disciplined (see sidebar on disparate consequences for dress code
violations). For example, our prior work showed that boys, Black students,
and students with disabilities are disproportionately disciplined across
discipline types, including exclusionary discipline. Other studies also
show Black and Hispanic students are more likely to receive harsher
32
See GAO, K-12 Education: Discipline Disparities for Black Students, Boys, and Students
with Disabilities, GAO-18-258 (Washington, D.C.: Mar 22, 2018).
Letter
Page 32 GAO-23-105348 K-12 Education
school discipline than their counterparts for the same violation.
33
One
study found that Black students were seven times more likely to receive
exclusionary discipline than their White peers.
34
District officials and
national organizations we spoke with echoed these findings and raised
concerns that, overall, dress codes can exacerbate disparities in school
discipline for Black students.
33
For examples of studies on the disproportionate impact for Black and Hispanic students,
see Matthew C. Fadus, Emilio A. Valdez, Brittany E. Bryant, Alexis M. Garcia, Brian
Neelon, Rachel L. Tomko, and Lindsay Squeglia, “Racial Disparities in Elementary School
Disciplinary Actions: Findings From the ABCD Study,” Journal of the American Academy
of Child & Adolescent Psychiatry. 60, no. 8. (August 2021): 998-1009 and Russell J.
Skiba, Robert H. Horner, Choong-Geun Chung, M. Karega Rausch, Seth L. May, and Tary
Tobin, “Race is Not Neutral: A National Investigation of African American and Latino
Disproportionality in School Discipline,” School Psychology Review, 40, no. 1 (2011): 85-
107.
34
See Aydin Bal, Jennifer Betters-Bubon, and Rachel E. Fish, “A Multilevel Analysis of
Statewide Disproportionality in Exclusionary Discipline and the Identification of Emotional
Disturbance,” Education and Urban Society, 51, no. 2 (2017): 247-268.
Education Case:
Disparate Consequences for Dress Code
and Other Discipline in a California District
In 2022, Education’s Office for Civil Rights
(OCR) found evidence that a California school
district engaged in disparate treatment based
on race in violation of Title VI by disciplining
Black students more frequently and more
harshly than similarly situated White students.
For example, an eighth-grade Black student
was referred for creating a hostile education
environment for wearing his pants low
(sagging) and refusing to pull his pants up
after repeated warnings. It was his first
discipline incident of the school year and he
received a one-day out-of-school suspension.
By contrast, a White eighth-grade student at
the same school was referred for obscenity for
sagging his pants in class after prior warnings.
In student interviews, Black students at one
district school reported to OCR that an
Assistant Principal followed them and treated
them differently from students of other racial
groups, including with respect to dress code
violations. Similarly, at least three students at
another school mentioned concerns about
how the school disproportionately applies the
dress code to Black girls.
To address the violations OCR found and to
ensure non-discrimination in student
discipline, the district entered into a resolution
agreement and committed to conduct a root
cause analysis to examine the causes of
racial disparities in its student discipline and
develop and implement a corresponding
Corrective Action Plan, among other
measures.
Source: GAO summary of Department of Education
document. | GAO-23-105348
Letter
Page 33 GAO-23-105348 K-12 Education
Further, officials from national organizations and districts we spoke with
raised concerns that Black girls may be particularly vulnerable to harm
from dress code enforcement. For example, they noted that dress codes
prohibiting hair coverings and hairstyles could be enforced more often
against Black girls. One study showed that Black girls are disciplined
primarily for less serious and more subjective offenses, such as disruptive
behavior, dress code violations, disobedience, and aggressive behavior.
35
This study also showed that Black girls are three times more likely than
White girls to be referred to the school office. Officials we interviewed at
national organizations also noted that Black girls may be perceived as
wearing more “revealing” or “distracting” clothing because research
shows that Black girls are mistakenly perceived to be older or more
mature.
Although Education does not have guidance that addresses disparities in
discipline enforcement, the agency recently signaled interest in issuing
resources to assist K-12 schools with improving school climate and safety
in the context of discipline. OCR has a goal, consistent with the civil rights
laws the agency enforces, to ensure equal access to education programs
and activities. In June 2021, Education requested information from the
public on discipline issues, including issues related to dress codes by July
23, 2021.
36
However, as of September 2022, it had not issued resources
on these topics. Federal information on how dress codes and other
discipline policies can be enforced in an equitable manner, and address
potential disproportionality, could help support and build schools’ capacity
to promote positive, inclusive, safe, and supportive school climates for
everyone.
35
Edward Morris and Brea Perry, “Girls Behaving Badly? Race, Gender, and Subjective
Evaluation in the Discipline of African American Girls,” Sociology of Education, 90, no. 2
(April 2017): 127-148.
36
Request for Information Regarding the Nondiscriminatory Administration of School
Discipline, 86 Fed. Reg. 30,449, 30,452 (June 8, 2021). On July 19, 2022, Education
issued guidance documents concerning the rights of students with disabilities in
connection with student discipline. These documents do not address dress code
discipline. See, U.S. Department of Education, Office for Civil Rights, Supporting Students
with Disabilities and Avoiding the Discriminatory Use of Discipline under Section 504 of
the Rehabilitation Act of 1973 (July 2022),
https://www2.ed.gov/about/offices/list/ocr/docs/504-discipline-guidance.pdf and U.S.
Department of Education, Office of Special Education and Rehabilitative Services,
Questions and Answers: Addressing the Needs of Children with Disabilities and IDEA’s
Discipline Provisions, (July 2022),
https://sites.ed.gov/idea/files/qa-addressing-the-needs-of-children-with-disabilities-and-ide
a-discipline-provisions.pdf.
Letter
Page 34 GAO-23-105348 K-12 Education
DressCodeEnforcementIsLinkedtoInformalRemovals
andNonExclusionaryDisciplinebutLittleIsKnownabout
TheirPrevalenceandEffects
In addition to documented suspensions and expulsions, researchers we
spoke with noted that dress code violations can also lead to sending
students home without formally suspending them. Such practices—often
referred to as “informal removals”—are not captured in Education’s
discipline data (see sidebar). In our generalizable review of dress codes,
we measured indicators of informal removals and estimate that 44
percent of all districts nationwide have policies that describe discipline
that removes students from the learning environment, but do not call
these removals suspensions or expulsions. These dress codes included
descriptions of removing students from class or sending them home (see
fig. 13).
Q&A: What are informal removals?
An “informal removal,” refers to an
administrative removal of a child from the
learning environment for a period of time
without documenting the removal as a
suspension or expulsion.
Informal removals can include being removed
from class or being “sent home” by the school.
For students with disabilities, this could
involve shortened school days or mandatory
homebound placement with little or no
education, or other methods.
Source: GAO summary of reports on informal removals;
Cavan for Adobe/stock.adobe.com. | GAO-23-105348
Letter
Page 35 GAO-23-105348 K-12 Education
Figure 13: Dress Code Excerpts: Examples of Removing Students from Class
Accessible Data for Figure 13: Dress Code Excerpts: Examples of Removing Students from Class
· Large suburban district: Students who are in violation of this rule will be sent to Student Services and
sent home with an unexcused absence for the day.
· Rural district: Students cited for dress code violations may be sent home for the day resulting in them
receiving an unexcused absence.
· Rural district: Students who miss class due to a dress code violation will be considered unexcused.
Attendance procedures will be followed. [For the] 6th Offense: Referral to court system for habitual
truancy charges.
· Large suburban district: If a student is out of dress code, guardians will be called and the student will
remain in the office until they can change. No exceptions.
· Rural district: Students who do not comply with the dress regulations will be sent home to change...
Time from school will be unexcused and makeup work will not be accepted.
Source: GAO review of dress codes. | GAO-23-105348
Given the wide recognition that exclusionary discipline policies can cause
negative educational and other outcomes, schools and districts face
pressure to lower their suspension and expulsion rates. In 2021,
Education noted that from the 2015-16 to 2017-18 school years, there
was an overall 2 percent decline in the use of exclusionary discipline
practices in public schools.
37
However, researchers we interviewed
cautioned that, to interpret these discipline trends (and explore possible
underreporting), it is important to examine trends in absenteeism to
37
U.S. Department of Education, Office for Civil Rights, An Overview of Exclusionary
Discipline Practices in Public Schools for the 2017-18 School Year (June 2021), available
at https://www2.ed.gov/about/offices/list/ocr/docs/crdc-exclusionary-school-discipline.pdf.
Letter
Page 36 GAO-23-105348 K-12 Education
capture instances where students are excluded from school in a way that
does not show up in discipline data.
38
Education has noted challenges related to informal removals in several
agency documents. In 2022, OCR found that a California district was
sending students home from school without officially suspending them.
This practice, and others, disproportionately harmed Black students. In
July 2022 guidance and an accompanying fact sheet, Education noted
that schools sometimes exclude students without triggering the school’s
formal disciplinary procedures, such as by requiring that a child be picked
up early from school.
39
However, these guidance documents do not describe the frequency of
these practices, which is needed to understand the effects on students
and student learning. Understanding the prevalence and impacts of the
full range of ways children are removed from the learning environment—
including informal removals—is important, because, as Education has
noted, school discipline can have significant impacts on student outcomes
and the learning environments of schools. Further, one national
stakeholder organization reported that “informal removals not only hurt
children academically and emotionally, but also harm their families,
communities, and society at large. Parents are often forced to scramble to
make arrangements in the middle of the workday because their child is
suddenly ‘out of school’.” They noted that informal removals, like other
practices that leave children unsupervised, are potentially detrimental to
children’s development.
40
Education relies on information about exclusionary discipline to inform its
efforts to provide resources on the equitable enforcement of discipline.
For example, as part of its civil rights investigations, OCR identified
instances of insufficient record keeping related to informal removals and
38
We attempted to explore the association between suspension rates and chronic
absenteeism over time, but the CRDC’s definition for chronic absenteeism changed in
school year 2017-18, making trend analysis unreliable.
39
See OCR’s 2022 guidance concerning the rights of students with disabilities in
connection with student discipline
https://www2.ed.gov/about/offices/list/ocr/docs/504-discipline-guidance.pdf and
https://www2.ed.gov/about/offices/list/ocr/docs/504-discipline-factsheet.pdf.
40
National Disability Rights Network, Out from the Shadows: Informal Removal of Children
with Disabilities from Public Schools (January 2022),
https://www.ndrn.org/resource/out-from-the-shadows.
Letter
Page 37 GAO-23-105348 K-12 Education
has sought to address these concerns. More generally, Education has
called for accurate records on the basis for removing a student and the
amount of time the student was removed from the learning environment,
noting in 2022 guidance on supporting students with disabilities that the
information is needed to determine appropriate provisions for students
with disabilities.
41
However, our findings raise concerns that students
experience exclusionary discipline more frequently than federal data
collections indicate.
In addition to concerns about exclusionary discipline, officials from
districts and national organizations raised concerns about less severe
types of discipline associated with dress code violations that do not
remove students from class (non-exclusionary discipline). This can
include verbal reprimands, requiring students to wear clothing that is not
their own, calling parents, after-school detention, and taking away
privileges and extracurricular activities. More than three-quarters of dress
codes (an estimated 77 percent) contain non-exclusionary disciplinary
options for enforcing dress codes. For our analysis of non-exclusionary
discipline options, see figure 14.
Figure 14: Non-Exclusionary Discipline Cited in Dress Codes Violations, by Estimated Percent of Districts
41
See OCR’s 2022 guidance concerning the rights of students with disabilities in
connection with student discipline, https://www2.ed.gov/about/offices/list/ocr/docs/504-
discipline-guidance.pdf and https://www2.ed.gov/about/offices/list/ocr/docs/504-discipline-
factsheet.pdf.
Letter
Page 38 GAO-23-105348 K-12 Education
Accessible Data for Figure 14: Non-Exclusionary Discipline Cited in Dress Codes Violations, by Estimated Percent of Districts
Estimated percentage of district dress codes
Category
Lower bound
Estimate
Upper bound
Requiring students change clothes
55.6
62.7
69.7
Escalating punishments for repeated offenses or refusal to comply
42.6
49.8
56.9
Calling parents/guardians
41.6
48.8
56
Detention
12.3
17
22.6
Requiring students wear clothing other than their own
7.4
11.5
16.9
Source: GAO review of district dress codes. | GAO-23-105348
Note: Using a 95 percent confidence interval, the margin of error for each school group is within +/- 8
percentage points.
Although a large body of research exists on exclusionary and other
severe discipline (such as corporal punishment), less is known about the
short- and long-term impact of non-exclusionary discipline on students,
for example, in terms of school engagement.
42
Education has noted that
punitive—including non-exclusionary—discipline can create a more
negative environment for all students in a school.
Education has a mission to promote student achievement and preparation
for global competitiveness, which it aims to support in a number of ways,
including collecting and disseminating data on America’s schools and
supporting evidence-based research on the effectiveness of education
programs and practices. To this end, Education’s Regional Educational
Laboratory (REL) Northwest had an “Equity in School Discipline”
Collaborative from 2017 through 2021. Its goal was to build practitioners’
capacity to use data and evidence to improve equitable school discipline
policies and practices.
43
Education officials also told us that REL West
recently entered into a partnership with San Francisco Unified School
42
In fact, one of the few studies we identified on non-exclusionary discipline recommended
further research on both exclusionary and non-exclusionary discipline. For example, see
V. Nishioka, B. Merrill, and H. Hanson, Changes in exclusionary and nonexclusionary
discipline in grades K–5 following state policy reform in Oregon, REL 2021–061
(Washington, D.C.: Department of Education, Institute of Education Sciences, National
Center for Education Evaluation and Regional Assistance, Regional Educational
Laboratory Northwest (2021), http://ies.ed.gov/ncee/edlabs.
43
RELs work in partnership with states and districts to (1) conduct original high quality
research, (2) provide training, coaching, and technical support, and (3) disseminate high
quality research findings to better understand their data, including research studies that
examine issues of equity and disproportionality.
Letter
Page 39 GAO-23-105348 K-12 Education
District focused on improving discipline outcomes for Black students in
the district. They noted that this partnership formed in response to
continuing inequitable patterns of student outcomes, including rates of
exclusionary discipline.
However, these efforts do not yet include research on the emerging issue
of informal removals, and Education officials noted that more research is
needed on non-exclusionary discipline. Education’s 2020 Data Strategy
calls for rigorous evaluations to identify effective programs, policies, and
practices, and guide educational institutions in their efforts to improve
student learning and outcomes.
44
Better information on the effects of non-
exclusionary discipline in schools, including dress code discipline, could
help achieve Education’s goal to use data to drive better operational
decision-making, foster educational excellence, and ensure equal access.
Conclusions
The need for school dress codes is often tied to health and safety, as
schools continue to grapple with how best to keep school communities
safe. However, researchers and officials from national organizations have
noted that some dress codes may create a less equitable and safe
environment for some students, especially girls, Black students, and
LGBTQI+ students. Specifically, common aspects of dress code
policies—such as taking measurements of students’ bodies and
clothing—may make school less safe for girls, in particular. Although
Education offers resources on ways schools can support racial, cultural,
and gender equity in general, and on school safety and climate, these
resources do not include information or examples about dress codes.
Given the prevalence of dress codes in public schools and the negative
impact poorly designed dress codes can have on students, Education has
an opportunity to further its goal of promoting safe, supportive learning
environments for all students by providing schools and districts with
information on designing equitable and safe dress codes.
There is wide recognition that exclusionary discipline, such as in-school
and out-of-school suspensions, and the resulting lost instructional time,
are associated with significant, negative educational and other long-term
44
Key evidence may come from a variety of sources, including data internal to Education,
data sourced from states and other grantees, and data from other agencies and
institutions. Department of Education, U.S. Department of Education Data Strategy
(December 2020).
Letter
Page 40 GAO-23-105348 K-12 Education
impacts for students. Schools that enforce strict dress codes have higher
rates of exclusionary discipline, which research has shown
disproportionately affects certain students. In addition, schools may
remove students “informally” from the classroom or use different types of
non-exclusionary discipline to enforce dress code violations, but these
measures are often unreported. As a result, the prevalence and impact of
these discipline practices are largely unknown. Education is uniquely
positioned to collect nationwide information on informal removals and
non-exclusionary discipline and inform future data collection and research
that captures a broader range of disciplinary actions and their effect on
student engagement and well-being. Accurate records and research on
disciplinary actions, including less severe types of discipline, are critical
for ensuring all students have equal access to educational opportunity, a
central component of Education’s mission.
RecommendationsforExecutiveAction
We are making the following four recommendations to Education:
The Secretary of Education should provide resources to help districts and
schools design equitable dress codes to promote a supportive and
inclusive learning environment. (Recommendation 1)
The Secretary of Education should include dress code information in
existing resources on safe and supportive schools. This information could
include examples of dress codes that safeguard students’ privacy and
body autonomy. (Recommendation 2)
The Secretary of Education should provide resources for states, school
districts, and schools on the equitable enforcement of discipline, including
dress code discipline. These resources should include information that
helps states, school districts, and schools address potential disparities
and disproportionality in dress code enforcement, as appropriate.
(Recommendation 3)
The Secretary of Education should collect information on the prevalence
and effects of informal removals and non-exclusionary discipline and
disseminate this information to states, school districts, and schools.
(Recommendation 4)
Letter
Page 41 GAO-23-105348 K-12 Education
AgencyCommentsandOurEvaluation
We provided a draft of this report to the Department of Education for
review and comment. Education provided technical comments, which we
incorporated as appropriate. In its formal comments, which are
reproduced in appendix VI, Education described the steps it is planning to
take to implement our four recommendations. With respect to our fourth
recommendation, which relates to collecting information on and assessing
the effect of informal removals and non-exclusionary discipline, Education
said that through a formal comment process, it is soliciting specific input
from the public on questions related to informal removals, and that the
responses will inform OCR’s proposed information collection request for
the CRDC’s 2025-26 school year collection. However, it also said it does
not have mechanisms for collecting information on the effects of these
practices. Education also said that IES is authorized to evaluate federal
education programs but that the department does not have a discrete,
evaluable program that addresses informal removals or non-exclusionary
discipline. Further, Education explained that while the agency does not
have the authority to direct RELs to conduct research on informal
removals and non-exclusionary discipline, the RELs could themselves
conduct such research if a REL’s stakeholders prioritize it.
We appreciate Education’s efforts to collect rigorous data on the subject,
and we understand the challenges of determining the effects of informal
removals and non-exclusionary discipline. We encourage Education to
think creatively about ways to collect and share information, within its
existing authorities, about the effects of informal removals and
exclusionary discipline. For example, it could leverage discussions with
stakeholders, working groups, or exploratory committees as a first step
toward collecting and disseminating information on the effects of informal
removals and non-exclusionary discipline.
We are sending copies of this report to the appropriate congressional
committees, the Secretary of Education, and other interested parties. In
addition, the report is available at no charge on the GAO website at
https://www.gao.gov.
Letter
Page 42 GAO-23-105348 K-12 Education
If you or your staff have any questions about this report, please contact
me at (617) 788-0580 or [email protected]. Contact points for our Offices
of Congressional Relations and Public Affairs may be found on the last
page of this report. GAO staff who made key contributions to this report
are listed in appendix V.
Jacqueline M. Nowicki, Director
Education, Workforce, and Income Security Issues
Letter
Page 43 GAO-23-105348 K-12 Education
List of Addressees
The Honorable Patty Murray
Chair
The Honorable Roy Blunt
Ranking Member
Subcommittee on Labor, Health and Human Services, Education, and
Related Agencies
Committee on Appropriations
United States Senate
The Honorable Rosa L. DeLauro
Chair
The Honorable Tom Cole
Ranking Member
Subcommittee on Labor, Health and Human Services, Education, and
Related Agencies
Committee on Appropriations
House of Representatives
The Honorable Robert C. “Bobby” Scott
Chairman
Committee on Education and Labor
House of Representatives
The Honorable A. Donald McEachin
House of Representatives
Appendix I: Objectives, Scope, and
Methodology
Page 44 GAO-23-105348 K-12 Education
AppendixI:Objectives,Scope,
andMethodology
Overview
This report examines (1) the characteristics of K-12 dress codes across
school districts nationwide, and how Education supports the design of
equitable and safe dress codes; and (2) what is known about the
enforcement of dress codes, and how Education supports equitable dress
code enforcement.
To conduct this work, we generated a nationally generalizable, stratified
random sample of school districts and analyzed publicly available dress
code information from these districts. Using this sample, we estimated the
prevalence of characteristics of dress codes in the United States. We also
conducted descriptive and regression analyses on the most recent years
of data available at the time of our review from Education’s School Survey
on Crime and Safety (school survey), Civil Rights Data Collection
(CRDC), and Common Core of Data (CCD). We assessed the reliability of
these data by reviewing existing documentation about the data and
performing electronic testing on required data elements. We determined
the data were sufficiently reliable for our purposes. Our descriptive
analyses focused on variables related to the prevalence of strict dress
code and uniform requirements and school demographics. See appendix
II for detailed information on our regression analyses. Additionally, we
identified and reviewed relevant studies on dress code discipline
outcomes in K-12 public schools.
To inform all aspects of our work, we interviewed academic researchers,
officials from national organizations, and federal agency officials from
Education. We also interviewed officials from three selected school
districts that recently revised their dress codes. In addition, we collected
anecdotal perspectives from families on dress codes and uniforms
through social media and a national organization. We also selected
examples of dress code enforcement in media reports published from
April 2018 to June 2022. Finally, we reviewed relevant federal agency
documentation, laws, and regulations. We also reviewed Education’s
Office for Civil Rights’ publicly available pending and resolved complaints
and cases related to dress code discipline in schools. The following
Appendix I: Objectives, Scope, and
Methodology
Page 45 GAO-23-105348 K-12 Education
sections contain detailed information about the scope and methodology
for this report.
ReviewofDressCodes
To describe the characteristics of K-12 dress codes nationwide, we
selected a nationally generalizable, stratified random sample of U.S. local
educational agencies (LEA), which we refer to as school districts
throughout the report. We used CCD for school year 2020-21 to select a
generalizable sample of public school districts using a sampling strategy
to ensure certain subpopulations of students were represented.
We based our sample on the LEA Universe database from the 2020-21
CCD and defined our sampling frame (the list of LEAs from which we
drew the sample) as follows. We included all operating LEAs:
· located in the District of Columbia, 50 states, or territories;
· with one or more schools;
· not closed according to the 2020-21 school year CCD, the most
recent data available at the time of our review; and
· offering any grades besides pre-K and ungraded.
1
From this sampling frame, we selected a stratified, random sample of 236
LEAs. Specifically, we stratified the sample frame of 17,600 LEAs into 13
mutually exclusive strata based on LEA size (largest 10), racial/ethnic
composition classification, urban classification, and charter status. We
selected the largest 10 LEAs, according to student count, with certainty
and determined the minimum sample size needed to achieve estimates of
plus or minus 10 percentage points or fewer, at the 95 percent confidence
level, within the following groups (see table 1). The original frame
included around 10,000 districts that were classified as state and federal
agency-run districts and agencies/organizations that provide educational
services and support to other schools/districts. These were outside of our
population of interest because they are not K-12 public districts and
instead provide support to other schools. Of these, seven districts were
part of our initial sample and were replaced by the next randomly selected
district within the same strata as the originally sampled district.
1
We excluded school districts classified in the CCD as supervisory union administrative
centers or federally or state operated institutions charged with providing elementary and
secondary instruction or services.
Appendix I: Objectives, Scope, and
Methodology
Page 46 GAO-23-105348 K-12 Education
Table 1: Population, Sample, and Completed Review Counts for Our In Scope
School Districts
Strata
Population
count
Sample
size
Completed
reviews
1. Top 10 student enrollment
10
10
10
2. Majority White (with unknown), Urban,
Non-Charter
285
5
5
3. Majority White (with unknown), Urban,
Charter
451
5
3
4. Majority White (with unknown), Suburban,
Non-Charter
2106
18
20
5. Majority White (with unknown), Suburban,
Charter
404
5
5
6. Majority White (with unknown),
Town/Rural (with unknown), Non-Charter
7826
65
73
7. Majority White (with unknown),
Town/Rural (with unknown), Charter
467
5
5
8. Majority Non-White, Urban, Non-Charter
577
10
12
9. Majority Non-White, Urban, Charter
1958
31
35
10. Majority Non-White, Suburban, Non-
Charter
1032
17
19
11. Majority Non-White, Suburban, Charter
579
10
12
12. Majority Non-White, Town/Rural (with
unknown), Non-Charter
1690
27
30
13. Majority Non-White, Town/Rural (with
unknown), Charter
215
5
5
Total
17600
213
236
Source: GAO analysis of the Department of Education’s Common Core Data for school year 2020-21. | GAO-23-105348
We allocated the sample across urban classification and charter status
LEA strata in proportion to the representation of each strata within its
race/ethnicity group.
We increased the sample size within each non-certainty stratum in order
to achieve the necessary number of completed reviews for our desired
precision level.
2
In addition, any strata that did not have a minimum
number of five observations was increased to five. We had an unweighted
and weighted response rate of approximately 92 percent, with 218 out of
the 236 sampled districts having a dress code policy available on the
district website. Because of the high response rate, and small
2
Based on our pretesting, we estimated a 90 percent availability rate of dress codes
based on our pre-testing of the document review tool with randomly selected districts.
Appendix I: Objectives, Scope, and
Methodology
Page 47 GAO-23-105348 K-12 Education
nonresponse count (18), carrying out a nonresponse bias analysis, such
as by assessing the association between nonresponse and other
variables in our frame, was not possible. We applied a within strata
nonresponse adjustment to the basic sampling weight to create an
analysis weight that represented the initial in scope population. We
analyzed our sample accounting for the complex sample design by using
survey software, specifying the strata, and using the analysis weights.
Because our sample is only one sample that might have been selected
using our methodology, we express the precision of our sample’s results
as a 95 percent confidence interval. This is the interval that would contain
the actual population value for 95 percent of the samples we could have
drawn. As a result, we are 95 percent confident that each of the
confidence intervals in this report will include the true values in the study
population.
To conduct our systematic review of selected district dress codes we
developed a structured data collection instrument. The data collection
instrument included topics such as the purpose of dress codes, prohibited
clothing items, rules about accessories or hairstyles, rules on clothing fit
and coverage, and consequences for dress code violations. We retrieved
dress code policies from school district websites. If no standalone dress
code policy was available, we reviewed any sections related to dress
codes in the district’s student or parent handbooks or codes of conduct, if
available. If a selected district had different versions of their dress code
by school level (e.g., elementary, middle school, and high school), we
reviewed the dress code for the highest school level available.
Each dress code policy was reviewed by two analysts. The two analysts
discussed any differences of opinion and reached agreement on the most
accurate representation of each dress code policy. In addition, to
minimize non-sampling error, we pre-tested draft versions of the data
collection instrument to ensure questions were clear and captured the
elements common in K-12 public school dress codes.
AnalysisoftheSchoolSurveyonCrimeandSafety
The School Survey on Crime and Safety—referred to in the body of this
report as the school survey—is a nationally representative survey of
principals in K-12 public schools conducted by the National Center for
Education Statistics (NCES). The survey samples approximately 4,800
U.S. public school principals or other administrators, and is large enough
to provide national estimates of all public schools while taking into
Appendix I: Objectives, Scope, and
Methodology
Page 48 GAO-23-105348 K-12 Education
account a number of factors such as level of instruction, student
enrollment size, and degree of urbanization, according to NCES. The
school survey generally covers nine topics: school practices and
programs, parent and community involvement at school, school security
staff, school mental health services, staff training and practices,
limitations on crime prevention, incidents, disciplinary problems and
actions, and school characteristics.
Our analyses primarily focused on two questions in the school survey:
“Was it a practice of your school to enforce a strict dress code?” and
“Was it a practice of your school to require students to wear uniforms?”
The school survey does not define “a strict dress code” or provide
instructions specific to these questions so it is possible that respondents
interpreted these questions differently, according to Education officials. In
addition, the school survey data are self-reported by principals or other
administrators, and consequently, as is generally true with self-reported
data, there is potential for misreporting of information. Our analysis was
conducted using the restricted-use data file of the school survey for
school years 2015-16 and 2017-18. According to Education officials, this
was the most recent data available at the time of our analysis.
We conducted a descriptive analysis of the school survey to determine
how school policies on dress codes and uniforms vary by school
characteristics such as a schools’ locale, size, level (e.g., elementary or
high school), student demographics, and percent of students who were
eligible for free or reduced-price lunch.
We also conducted generalized linear regressions using the 2017-18
school survey data to explore associations between enforcing a strict
dress code and school-level and student characteristics while controlling
for other factors.
3
We conducted a similar analysis for schools that have
uniforms. Please see the technical appendix II for detailed regression
specifications.
3
Specifically, we used multivariate logistic regression because the outcome variable is
binary (taking value 1 if the school enforces a strict dress code and 0 otherwise). A logistic
regression model provides an estimated odds ratio of an event occurring, such as whether
a school characteristic is associated with higher or lower odds of strict dress code
enforcement, holding other factors constant.
Appendix I: Objectives, Scope, and
Methodology
Page 49 GAO-23-105348 K-12 Education
AnalysisoftheCivilRightsDataCollectionandthe
SchoolSurveyonCrimeandSafety
The Civil Rights Data Collection (CRDC) is a biennial survey that is
mandatory for nearly every public school and school district (pre-K – 12th
grade) in the 50 states, D.C., and Puerto Rico, and is conducted by
Education’s Office for Civil Rights. The data contain information about
school characteristics and about programs, services, and outcomes for
students. Most student data are disaggregated by race/ethnicity, sex,
English Learners, and disability status. Please see the technical appendix
II for more information.
We conducted generalized linear regressions using the 2015-16 and
2017-18 school survey data merged to data from the 2015-16 and 2017-
18 Civil Rights Data Collection (CRDC) and Common Core Data (CCD) to
explore associations between selected school-level policies and
characteristics and rates of incidents of exclusionary disciplinary
measures such as the percentage of students suspended, while
controlling for other factors.
4
Such a model allowed us to test the
association between disciplinary incidents and selected school-level and
student policies and characteristics, such as enforcing a strict dress code,
while holding other factors constant (such as school type, locality, and
student demographics). We conducted a separate regression for each of
the specific types of exclusionary disciplinary incidents, as well as having
policies on student clothing such as enforcing a strict dress code or
requiring a school uniform. Again, please see the technical appendix II for
detailed regression information.
TargetedLiteratureReview
To provide context for our data analysis on the enforcement of dress
codes in K-12 schools, we identified and reviewed relevant studies on
student outcomes associated with K-12 dress code disciplinary actions.
To identify these studies, we conducted searches of several social
science databases: Scopus, ProQuest, and EBSCO for keywords such as
“dress code” and “discipline.” We also asked the academic researchers
4
We used a Poisson generalized linear regression for this analysis because the data on
outcomes (e.g., number of students having in-school suspension) represent counts, and
therefore are not appropriate for a traditional normal linear model. In addition, we used a
negative binomial regression instead of a Poisson regression because negative binomial
models are appropriate for count analyses with observed over-dispersion (i.e., when the
variance of the count variable is much larger than the mean of that variable).
Appendix I: Objectives, Scope, and
Methodology
Page 50 GAO-23-105348 K-12 Education
and national organizations we interviewed to recommend studies. We
limited results to studies published from August 2011 to June 2022.
Relevant studies met the following criteria: (1) empirical research using
student-level data from public schools in the United States, (2) related to
student outcomes associated with K-12 dress code disciplinary actions,
(3) used rigorous statistical methods, and (4) were published in a peer-
reviewed journal. These studies were used to provide additional
information and context related to dress code discipline, in part because
there is no national data on discipline infractions.
InterviewswithSelectedSchoolDistrictOfficialsand
NationalOrganizations
To obtain information on how K-12 school districts have revised dress
codes, we selected three districts with which to conduct virtual site visits.
We identified districts that recently revised their student dress codes
through internet searches using terms such as “revised school dress
code,” “new school dress code,” and “new school uniform policy.” We
defined a policy as recently revised if it was updated within the last 5
school years. We also looked at dress code policies in school districts
with disciplinary incidents that gained news coverage, identified through
internet searches and interviews with national organizations, and in
neighboring districts that may have also felt pressure to revise dress
codes in response to local news coverage. Because this method relied on
policy revisions gaining news coverage, we did not identify all districts
that recently revised their dress codes.
From the districts we identified, we selected districts for variation
according to the following criteria: type of revision (i.e., making dress
code less strict or moving to uniforms), region of the country, and the
number of students enrolled in the district. We reviewed revision policy
documents from each selected district and conducted interviews with
district officials on the reasons for the revision, the revision process,
challenges, and outcomes of the new policy.
In addition, we interviewed officials from national organizations and
academic researchers. We selected organizations and researchers for
interviews based on their knowledge of relevant information and work
related to dress codes and discipline in K-12 public schools. From our
interviews, we gathered information on how dress code policies are
developed, enforced, any benefits or challenges related to dress code
Appendix I: Objectives, Scope, and
Methodology
Page 51 GAO-23-105348 K-12 Education
policies, and what support or assistance Education provides related to
dress codes.
FamilyQuestionnaire
We obtained non-generalizable information on families’ perspectives on
their children’s school dress codes and uniform policies. We circulated an
anonymous questionnaire through social media and a national
organization in March and April 2022. The questionnaire asked about
families’ experiences with school dress codes or uniforms and their views
on how it affected their child and their family, if at all. We received 47
completed questionnaires. The team analyzed responses for common
themes and selected illustrative examples that represent the range of
perspectives.
SelectedMediaReportsonDressCodeEnforcement
To provide illustrative examples of the types of incidents involving dress
code enforcement that have been reported in national, regional, or local
media reports, we created an internet search news alert for “school dress
code” which limited media reports to the United States. We reviewed
media reports identified through this search regularly between January
and June 2022 and compiled a list of relevant media reports. The team
also conducted internet searches using specific keywords to identify
incidents where students were disciplined for dress code violations and
are members of certain identity groups related to race, color, or national
origin; religion; and sex (including sexual orientation and gender identity).
In addition, we identified media reports from other background research.
We selected media reports from the compiled list to illustrate the variety
of dress code policies, enforcement strategies, and concerns identified by
our generalizable review of district dress codes, interviews with district
officials, and targeted literature review. We also analyzed the incidents
using sufficiency and relevancy criteria, including whether the media
report had information about the nature of the incident and where it
occurred. The media reports provide illustrative descriptions of individual
incidents and do not represent the experiences of all students and
schools. We did not assess whether the incidents could potentially
constitute unlawful discrimination under federal or state law.
Appendix I: Objectives, Scope, and
Methodology
Page 52 GAO-23-105348 K-12 Education
ReviewofRelevantFederalLaws,Regulations,
Programs,andDocumentsRelatedtoDressCode
Discipline
We reviewed relevant federal laws, regulations, programs, and
documents related to dress code discipline, informal removals, and
school climate and safety. We interviewed or sent questions to the
following offices at the Department of Education: Office for Civil Rights
(OCR), Office of Elementary and Secondary Education, Institute of
Education Sciences, and the Office of Special Education and
Rehabilitative Services. We also explored publicly available OCR cases
to find examples related to K-12 dress code discipline.
We conducted this performance audit from August 2021 to October 2022
in accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings and
conclusions based on our audit objectives.
Appendix II: Technical Appendix for
Regression Analyses
Page 53 GAO-23-105348 K-12 Education
AppendixII:TechnicalAppendix
forRegressionAnalyses
We conducted generalized linear regressions using the 2015-16 and
2017-18 School Survey on Crime and Safety (SSOCS) data merged to
data from the 2015-16 and 2017-18 Civil Rights Data Collection (CRDC)
and Common Core Data (CCD). Our analyses explored associations
between selected school-level policies and characteristics, and rates of
incidents of exclusionary disciplinary measures, such as the percentage
of students suspended, while controlling for other factors.
1
Similarly, we
conducted generalized linear regressions to explore associations
between selected student and school-level characteristics and policies
and whether or not a school enforces a strict dress code or has school
uniforms, while controlling for other factors. Such a model allowed us to
test the association between disciplinary incidents and selected student
and school-level characteristics and policies, such as enforcing a strict
dress code, while holding other factors constant (such as school type,
locality, and student demographics.) We conducted a separate regression
for each of the specific types of exclusionary disciplinary incidents, as well
as having policies on student clothing such as enforcing a strict dress
code or requiring a school uniform.
Typically, a generalized linear regression model is appropriate when the
model assumption of normality is not appropriate, as is the case with a
binary (yes/no) outcome for logistic regressions, or a count outcome for
Poisson regressions. A logistic regression model provides an estimated
odds ratio of an event occurring, such as whether a school characteristic
is associated with higher or lower odds of strict dress code enforcement,
holding other factors constant. A Poisson regression model provides an
estimated incidence rate ratio of an event, such as whether a school
characteristic is associated with higher or lower rates of exclusionary
discipline such as suspensions, holding other factors constant.
1
We used a Poisson generalized linear regression for this analysis because the data on
outcomes (e.g., number of students having in-school suspension) represent counts and
therefore are not appropriate for a traditional normal linear model. In addition, we used a
negative binomial regression instead of a Poisson regression because negative binomial
models are appropriate for count analyses with observed over-dispersion (i.e., when the
variance of the count variable is much larger than the mean of that variable).
Appendix II: Technical Appendix for
Regression Analyses
Page 54 GAO-23-105348 K-12 Education
For both the estimated odds ratio and estimated incidence rate ratio, a
value greater than one indicates a higher or positive association, and a
value less than one indicates lower or negative association, when the
factor is present. For example, an estimated odds ratio less than one
indicates lower odds of being suspended when a factor is present.
Additionally, one can quantify just how much more or less likely the
incidence is, according to the estimated model coefficients. For example,
an estimated incidence rate ratio of student suspension of 2.5 for a
majority female school would be associated with 2.5 times higher
incidence of suspensions relative to schools which are majority male,
holding all other variables in the model constant such as school type and
policies on student clothing. Given limitations of our models, including that
we must rely on observational data that did not come from an
experimental design which would allow for causal inference and the
unknown bias introduced from item nonresponse, we present a general
summary of associations by providing the direction, rather than an
estimated rate (incidence) of student discipline.
To obtain a better understanding of potential control variables and their
association with outcomes, and to identify potential controls used by
subject matter experts from studies using similar methodologies, a
literature review was performed. In particular, regression studies, which
were similar in scope to the engagement objectives, were reviewed and
summarized. Data from these regression studies represented a range of
school years. This information, in addition to prior GAO work, was used to
inform our final control variable selection.
Regression Analysis of the School Survey on Crime and Safety and Civil
Rights Data Collection
· We conducted generalized linear regressions using the 2015-16 and
2017-18 School Survey on Crime and Safety (SSOCS) data merged
to data from the 2015-16 and 2017-18 Civil Rights Data Collection
(CRDC) and Common Core Data (CCD). This is approximately the
same universe of about 2,700 and 2,100 U.S. school principals
surveyed from the 2017-18 and 2015-16 SSOCS survey years,
respectively. Because the data come from a nationally representative
survey, these data use sampling weights to allow for inferences to be
made about the larger population of schools from which the sample
units were drawn. For each year SSOCS covers a population of over
84,000 schools. In addition to incorporating sample weights, the data
contain replicate weights which were used in variance estimation to
account for the sample design. Because SSOCS data are from a
Appendix II: Technical Appendix for
Regression Analyses
Page 55 GAO-23-105348 K-12 Education
complex sample design, we analyzed the merged SSOCS, CRDC,
and CCD data using the analysis weights and sampling design
information in order to account for the complex sample design.
However some schools were excluded from our regression model if
data were not available in all sources. These schools are
approximately 3 percent of the total records for both years combined.
Of the schools in the 2017-18 SSOCS, 50 could not be matched; of
the schools in the 2015-16 SSOCS, 73 could not be matched.
All regression models are subject to limitations and for this model, the
limitations included:
· Data analyzed for these regression analyses were by school rather
than by student. Consequently, they are not able to describe the
association between our independent variables and a student’s
experience of disciplinary incidents, such as suspensions, while
controlling for characteristics of an individual student such as gender,
race or ethnicity, or grade level. Instead the school-level nature of the
SSOCS, CRDC, and CCD data limited this particular analysis of the
associations between school characteristics and policies as to
whether there was an increase, decrease, or no effect on the rate of
disciplinary incidents such as suspensions, controlling for other
characteristics of the entire school’s population, such as school type,
or percent of students who are female.
· Some variables which may be related to school characteristics and
policies and disciplinary incidents are not available in the data. For
example, in this context, it could be that a school’s average student
household income adjusted for family size that could be related to
students’ exposure to disciplinary incidents in schools, such as
suspensions.
· Results of our analyses are associational and do not imply a causal
relationship because, for example, SSOCS data are observational in
nature and were not gathered by a randomized controlled trial, where
students would be randomized to attend schools with certain
characteristics.
· Additionally, SSOCS data are subject to both sampling and non-
sampling error. While the analysis has accounted for sampling error,
survey data are also affected by non-sampling error which could occur
for many reasons, including a failure to sample a segment of the
population, inability to obtain information for all respondents in the
sample, inability or unwillingness of respondents to provide correct
information, mistakes by respondents, and errors made in the
Appendix II: Technical Appendix for
Regression Analyses
Page 56 GAO-23-105348 K-12 Education
collection or processing of data (such as imputation or data quality
checks).
· To determine the extent to which these non-matched SSOCS records
excluded from the analysis differed from the SSOCS schools which
were included, we conducted a unit bias analysis by comparing the
schools which were excluded and included and examining differences
across school characteristics and policies to identify potential sources
of bias. This was done separately for each survey year of SSOCS
data. The same school characteristics and policies from SSOCS data
which are used in the final regression model were examined in this
analysis. First, a logistic regression modeling relative propensity of
whether or not a school was matched while controlling for school
characteristics and policies was fit. To further investigate potential
bias, weighted frequency tables and estimates such as averages were
compared across the two response groups. Some estimates may be
subject to nonresponse bias that is not related to the observable
characteristics used to conduct the bias analysis. Because certain
school characteristics are not observed for non-matched records, the
exact amount of bias remaining in estimates cannot be known with
certainty and is likely to vary between estimates. Because of this,
findings are presented as positive or negative associations rather than
odds or relative risk ratios.
For the purposes of our analysis we created some composite or recoded
variables (see table 2). Table 3 lists the control, or independent variables
we included in our regression models of the rate of exclusionary discipline
actions. We conducted a separate regression for each of the three
exclusionary disciplinary measures at school, and separate regressions
were also run by demographic subgroup of students.
Appendix II: Technical Appendix for
Regression Analyses
Page 57 GAO-23-105348 K-12 Education
Table 2: Created Variables Used in the Regression Analysis of Department of Education’s School Survey on Crime and Safety
and Civil Rights Data Collection, School Years 2015-2016, and 2017-2018
GAO category
Variables from SSOCS, CRDC, or CCD
Recoded value(s)
School type
· Type of school (SSOCS C0564):
· Regular public school
· Charter school
· Has a magnet program for part of the school
· Exclusively a magnet school
· Other (specify)
· Regular public school
· Magnet school (exclusively or
partially)
· Charter or other school
Net transfers
· Transferred to (SSOCS C0570)
· Transferred from (SSOCS C0572)
· Transferred to (minus) Transferred
from (continuous)
School disorder
· Student racial ethnic tensions (SSOCS C0374)
· Student verbal abuse of teachers (SSOCS C0380)
· Widespread disorder in classrooms (SSOCS C0382)
· Student acts of disrespect for teachers (SSOCS C0384)
· Gang activities (SSOCS C0386)
· Regular (if at least one occurs daily or
weekly)
· Rare (if else at least one occurs
monthly or occasionally)
· Never (if all never occur)
Census Region
LEA state (CRDC)
· Northeast (New England and Middle
Atlantic)
· Midwest (East North Central and
West North Central)
· South (South Atlantic, East South
Central, and West South Central)
· West (Mountain and Pacific)
· Other outlying area
Source: GAO analysis. | GAO-23-105348
Appendix II: Technical Appendix for
Regression Analyses
Page 58 GAO-23-105348 K-12 Education
Table 3: Variables Included in Our Regression Models Using the Department of Education’s School Survey on Crime and
Safety and Civil Rights Data Collection, School Years 2015-2016, and 2017-2018
Control/Independent variables
Outcome/dependent Variables
Model specification
School characteristics: School disorder, school offers grade(s) 6
or above, net transfers, school Type
School staff: Number of full time equivalent security staff, student
to teacher ratio, student to counselor ratio, percentage of
teachers with less than 2 years of experience
Student characteristics: Percentage of students: in special
education program, in attendance each day, scoring below the
15th percentile on standardized tests, going to college, have free
or reduced price lunch, students with Individuals with Disabilities
Education Act (IDEA) disabilities, female students, Black female
students, Hispanic/Latino female students; majority non-white
school
School policies: school enforces a strict dress code, school
requires a uniform
School geography: Census region, locale
Year of SSOCS data
Number of: total students, female
students, male students, Black students,
Hispanic/Latino students, White students,
students with IDEA disabilities, students
without IDEA disabilities experiencing the
following disciplinary outcomes:
· Out of school suspension
· In school suspension
· Expulsion
Poisson/Negative
binomial regression
a
Source: GAO analysis. | GAO-23-105348
a
Because rates of students experiencing disciplinary actions are of interest, an exposure variable was
used representing the total population of students. For example, when modeling the number of
female students experiencing suspension the exposure is set to the total number of female students.
Appendix II: Technical Appendix for
Regression Analyses
Page 59 GAO-23-105348 K-12 Education
Table 4: Variables Included in Our Regression Model Using the Department of Education’s School Survey on Crime and
Safety and Civil Rights Data Collection, 2017-2018
Control/Independent Variables
Outcome/dependent variables
Model specification
School characteristics: school offers grade(s) 6 or above,
charter school, total enrollment
Student characteristics: Percentage of students: in special
education program, female students, Black students,
Hispanic students, other race (Asian, American Indian,
Native Hawaiian, Alaskan Native, two or more races)
School geography: Census region
Year of SSOCS data: 2017-18
Whether school reports enforcing a strict
dress code (Binary)
Whether school reports requiring a uniform
(Binary)
Logistic regression
Source: GAO analysis. | GAO-23-105348
Given the limitations of our model as described above, we present the
results of our regression models by describing the direction of the
associations. “Increase” means that a particular variable was significantly
associated with an increase in the rate of exclusionary discipline, such as
the rate of in-school suspensions, at the 0.05 level; “decrease” indicates a
decrease in the rate or odds, while holding all other variables in the model
constant. Insignificant indicates the variable is not significantly associated
with the given exclusionary discipline action at the 0.05 level. For
categorical variables in these tables, we provided the comparison school
characteristic in the “Effect of Variable” column. For example, the results
in these tables should be interpreted as a school reporting enforcement of
a strict dress code is more likely to report higher rates of students given
in-school suspension relative to schools which do not report enforcement
of a strict dress code, holding other factors constant, because the
association is positive.
It should be noted that interactions (i.e., where we combine indicators for
the number of full time security staff and school enforcement of a strict
dress code) should be interpreted differently than other variables. Though
an interaction may be associated with a “decreased” rate of discipline, it
does not necessarily imply that the group presented in the interaction was
significantly less likely to receive the disciplinary action because
interactions are interpreted relative to the main effect of each variable in
the interaction. Since the contribution for an interaction is relative, the
contribution of the main effects could interact to create a stronger effect
together relative to the effect alone. However, the contribution of the main
effects could outweigh the effect of the interaction, resulting in a positive
effect altogether despite a negative interaction.
Appendix III: Snapshot of Selected School
Districts That Revised Dress Codes
Page 60 GAO-23-105348 K-12 Education
AppendixIII:Snapshotof
SelectedSchoolDistrictsThat
RevisedDressCodes
Table 5: Snapshot of Selected School Districts That Revised Dress Codes
Category
Mid-western school district
Southeastern school district
West Coast school district
Reasons for revising
dress code
To enhance socio-economic
equity.
To enhance gender equity.
To enhance equity regardless of
gender, gender identity, sexual
orientation, race/ethnicity, body
type/size, religion, and personal style.
Revision strategies
· Focus groups with district
staff, teachers, families, and
students.
· Consulted with other
districts and the Council for
Great City Schools.
· Reviewed data on dress code
violations.
· Discussion groups with school
administrators, district staff,
teachers, families, and
students.
· Reviewed two “model” dress
codes.
· Focus groups and administrator
training.
· Surveyed staff,
parents/guardians, and students.
Key changes made
Adopted a uniform policy (e.g.,
pants must be tan, navy blue, or
black with polo shirts in navy
blue, black, or an approved
school color).
Removed gender based language
and simplified rules to only require
students to cover areas from chest
to mid-thigh.
Adopted gender-neutral dress code
(with considerations for nonbinary
students) and limited subjective
language and staff interpretation.
Key challenges
post-implementation
Many schools and students
opted-out of the uniform policy.
Challenges implementing rules on
head coverings and gang-related
attire equitably.
Staff push back on allowing
hats/hoods, saying it made it difficult
to see students’ faces and
earphones.
Reported results
No change in the number of
dress code violations after the
uniform policy was implemented.
No change in the number of dress
code violations after the new dress
code was implemented.
Dress code violations decreased
under the new dress code.
Source: GAO review of documents and interviews with district officials. | GAO-23-105348
Appendix IV: Comments from the U.S.
Department of Education
Page 61 GAO-23-105348 K-12 Education
AppendixIV:CommentsfromtheU.S.
DepartmentofEducation
Appendix IV: Comments from the U.S.
Department of Education
Page 62 GAO-23-105348 K-12 Education
Appendix IV: Comments from the U.S.
Department of Education
Page 63 GAO-23-105348 K-12 Education
Accessible Text for Appendix IV: Comments
from the U.S. Department of Education
Page 64 GAO-23-105348 K-12 Education
AccessibleTextforAppendixIV:
CommentsfromtheU.S.Department
ofEducation
October 12, 2022
Jacqueline M. Nowicki
Director, Education, Workforce,
and Income Security Issues
Government Accountability Office
441 G Street, N.W.
Washington, D.C. 20548
Dear Ms. Nowicki:
I write on behalf of the U.S. Department of Education (Department) in response to
the statements and recommendations made in the Government Accountability Office
(GAO) draft report, K-12 EDUCATION: Department of Education Should Provide
Information on Equity and Safety in School Dress Codes (GAO-23-105348). We
appreciate the opportunity to respond.
The draft report examines the characteristics of dress codes in K-12 public schools,
the enforcement of dress codes, and how the Department supports the design of
dress codes and equitable enforcement of them. The draft report includes four
recommendations to the Secretary of Education. This letter provides our responses
below.
Recommendation 1: The Secretary of Education should provide resources to help
districts and schools design equitable dress codes to promote a supportive and
inclusive learning environment.
Response: The Department will work with appropriate technical assistance centers to
highlight available high-quality resources and, as needed, develop new resources
around equitable dress codes as a strategy to help promote safe, supportive,
inclusive, and fair learning environments.
Recommendation 2: The Secretary of Education should include dress code
information in existing resources on safe and supportive schools. This information
Accessible Text for Appendix IV: Comments
from the U.S. Department of Education
Page 65 GAO-23-105348 K-12 Education
could include examples of dress codes that safeguard students’ privacy and body
autonomy.
Response: In coordination with the response to Recommendation 1, the Department
will review existing resources and, as needed, develop new resources on effective
dress code policies, practices, and programs that local educational agencies may
consider to promote safe, supportive, inclusive, and fair learning environments.
Recommendation 3: The Secretary of Education should provide resources for states,
school districts, and schools on the equitable enforcement of discipline, including
dress code discipline. These resources should include information that helps states,
school districts, and schools address potential disparities and disproportionality in
dress code enforcement, as appropriate.
Response: In coordination with the responses to Recommendations 1 and 2, the
Department will review opportunities to provide information to help school
communities address possible unlawful discrimination with respect to dress code
enforcement. The Department will also, as appropriate, include information on
equitable enforcement of discipline in any resources that may be developed
addressing discipline, including for dress and grooming codes.
Recommendation 4: The Secretary of Education should collect information on the
prevalence and effects of informal removals and non-exclusionary discipline and
disseminate this information to states, school districts, and schools.
Response: The Department is exploring the possibility of collecting data for future
Civil Rights Data Collections (CRDCs), after the 2021–22 and 2023–24 school year
CRDCs, on the prevalence of informal removals, such as removals from a class or a
school, to gauge inequities in the use of informal removals as student discipline.
Through the Paperwork Reduction Act comment process, the Office for Civil Rights
(OCR) is currently soliciting specific input from the public on questions, including:
what are the common types of informal removals?; how should OCR define informal
removals of students?; what specific data involving students who receive informal
removals should OCR collect?; and what data are school districts and schools
currently collecting regarding informal removals? The responses will inform OCR’s
proposed information collection request for the 2025-26 school year collection.
Because OCR collects only aggregate student administrative data (disaggregated by
race, sex, disability, etc.), OCR has no available mechanism to collect information on
the impact or effects of informal removals or non-exclusionary discipline on students
in terms of, for example, school engagement.
The Institute of Education Sciences (IES) is authorized to evaluate Federal education
programs administered by the Secretary (see sections 171(b) and 173 of the
Accessible Text for Appendix IV: Comments
from the U.S. Department of Education
Page 66 GAO-23-105348 K-12 Education
Education Sciences Reform Act of 2022; ESRA). At present, the Department does
not have a discrete, evaluable program that addresses informal removal or non-
exclusionary discipline, so the Department does not believe IES has the authority to
conduct an evaluation in order to address this recommendation. IES also administers
the Regional Educational Laboratory (REL) program, which conducts applied
research in partnership with state and local education agencies. The program of
research for each REL is designed to be responsive to needs identified by regional
stakeholders (see section 174 of ESRA). IES does not have authority to direct the
RELs to conduct applied research on this or any other topic. As noted in the draft
report, regional stakeholders have worked with their RELs to explore exclusionary
discipline in the past. To the extent that regional stakeholders prioritize informal
removal or non-exclusionary discipline as topics for research and appropriate data
are available, the incidence and outcomes associated with these practices could be
described at the state or local level.
Thank you for the opportunity to respond to the recommendations outlined in this
GAO draft report. We appreciate GAO’s work on this important issue.
Sincerely,
Catherine E. Lhamon
Assistant Secretary for Civil Rights
Appendix V: GAO Contact and Staff
Acknowledgments
Page 67 GAO-23-105348 K-12 Education
AppendixV:GAOContactand
StaffAcknowledgments
GAOContact
Jacqueline M. Nowicki, (617) 788-0580 or [email protected].
StaffAcknowledgments
In addition to the contact named above, Nagla’a El-Hodiri (Assistant
Director), Alison Grantham (Analyst-in-Charge), Abby Marcus, and
Jessica L. Yutzy made key contributions to this report. Also contributing to
this report were James Bennett, Elizabeth Calderon, Tonnyé Conner-
White, Sherri Doughty, Randi Hall, Abigail Loxton, Abinash Mohanty,
John Mingus, Jessica Orr, David Reed, Monica Savoy, Almeta Spencer,
Frances Tirado, Manuel Valverde, and Sonya Vartivarian.
(105348)
GAO’sMission
The Government Accountability Office, the audit, evaluation, and investigative
arm of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of the
federal government for the American people. GAO examines the use of public
funds; evaluates federal programs and policies; and provides analyses,
recommendations, and other assistance to help Congress make informed
oversight, policy, and funding decisions. GAO’s commitment to good government
is reflected in its core values of accountability, integrity, and reliability.
ObtainingCopiesofGAOReportsandTestimony
The fastest and easiest way to obtain copies of GAO documents at no cost is
through our website. Each weekday afternoon, GAO posts on its website newly
released reports, testimony, and correspondence. You can also subscribe to
GAO’s email updates to receive notification of newly posted products.
OrderbyPhone
The price of each GAO publication reflects GAO’s actual cost of production and
distribution and depends on the number of pages in the publication and whether
the publication is printed in color or black and white. Pricing and ordering
information is posted on GAO’s website, https://www.gao.gov/ordering.htm.
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card, MasterCard,
Visa, check, or money order. Call for additional information.
ConnectwithGAO
Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Subscribe to our RSS Feeds or Email Updates. Listen to our Podcasts.
Visit GAO on the web at https://www.gao.gov.
ToReportFraud,Waste,andAbuseinFederal
Programs
Contact FraudNet:
Website: https://www.gao.gov/about/what-gao-does/fraudnet
Automated answering system: (800) 424-5454 or (202) 512-7700
CongressionalRelations
A. Nicole Clowers, Managing Director, [email protected], (202) 512-4400, U.S.
Government Accountability Office, 441 G Street NW, Room 7125, Washington,
DC 20548
PublicAffairs
Chuck Young, Managing Director, [email protected], (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, DC 20548
StrategicPlanningandExternalLiaison
U.S. Government Accountability Office, 441 G Street NW, Room 7814,
Washington, DC 20548
