Examining the Provision of School-Based Athletics in Public High Schools Among Girls in the United States

Abstract

BACKGROUND:

In order to better understand some of the potential barriers to interscholastic sport for girls, we examine several school-level factors to assess how they influence the provision of sport for girls across an eight year time frame.

METHODS:

The data for this study comes from the Civil Rights Data Collection (CRDC) that collected school-level information for the majority of U.S. public schools between the 2009–10 and 2017–18 school years.

RESULTS:

Several important school-level factors were found to limit the provision of sport for girls and included factors related to the proportion of the student body who were eligible for free lunch and who were racial minorities. For instance, schools where 75% to 100% of students were eligible for free lunch typically offered 3 less sport teams for girls (b = −2.93, p<.001) when compared to schools where 0% to 24% of students were eligible for free lunch.

CONCLUSIONS:

This study provides evidence that schools with fewer resources and a higher proportion of disadvantaged students tend to have the fewest opportunities to participate in interscholastic sport, particularly among girls. Moreover, girls participation rates in sport still lag behind boys despite having similar numbers of sports and sport teams.

Participation in sport among adolescents has been found in many studies to be beneficial for both adolescents’ academic achievement and health.^{1–8, 41, 47–48} Research has also shown the long-term benefits of participation in interscholastic sports leads to better mental and physical health in adulthood ^42–44. Although some nationally representative studies have found that participation in certain types of sport are associated with health compromising behaviors or diminished academic outcomes,^9–13 the overwhelming majority of research has found that sport is a positive asset in the lives of youth.^{2,3, 47–48,50}

A variety of studies suggest a link between athletic participation and positive types of health related outcomes among adolescent athletes.^{6–8,14–21} The contributions of interscholastic sport to adolescent physical activity and other positive types of health related outcomes are both direct and indirect. Athletic participation itself constitutes a type of physical activity, and as such, it contributes to the overall physical activity of adolescents.^22,45 Factors like the length of a season, extent of practice time, and the expectations and demands of coaches directly influence overall physical activity rates. Indirectly, many adolescents construct their lifestyles and identities through and around sport.^23,48 In varying degrees, interscholastic sport provides a social environment in which coaches and parents can encourage adolescents to develop physical skills, take initiative to enhance their health and physical well-being, and integrate greater levels of physical activity into a healthy lifestyle.²⁴

Encouragingly, the provision of interscholastic sport has increased substantially over the past 40 years.²⁵ In fact, during the 2017–2018 school year, 7,980,886 adolescents participated in interscholastic sport at the high school level.²⁶ In other words, nearly half of all students who were enrolled in public high schools (i.e. total number enrolled = 15,100,000)²⁷ in the U.S. during the 2017–18 school year participated in interscholastic sport. While these numbers are reassuring, more effort needs to be directed on assessing these trends across time in order to determine if these opportunities are being equitably distributed to historically disadvantaged groups. Indeed, prior research that has examined these groups have found evidence that access to sport is limited for girls, minorities, and adolescents from low socioeconomic (SES) backgrounds.^8,28–29 For instance, research focused on the prior decade found that U.S. high schools provide fewer athletic opportunities for girls, particularly those who attend schools with a large proportion of students from low SES backgrounds.²⁹ Moreover, this same study found that public schools with a higher proportion of female students, minority students, and students from low SES backgrounds were more likely to have cut entire interscholastic sport programs between the 1999–2000 and 2009–10 school years.²⁹ Schools with fewer resources and a higher proportion of historically disadvantaged students tend to have the fewest opportunities to participate in interscholastic sport, and additionally, they are more likely to cut sports from their overall activities portfolio.^28–29 This is problematic since interscholastic sport may be the only opportunity for underserved populations to participate and learn about athletics given the decline in physical education in secondary schools,³⁰ the limited availability of recreational facilities in low-income communities,⁸ the financial constraints imposed by youth sport programs,³¹ and the subordinate nature of female-dominated sports.³²

Although empirical research has established links between involvement in sport and positive types of health related outcomes among youth, it is important to highlight that these links are stronger for girls when compared to boys with respect to certain developmental outcomes (e.g., taking AP courses in science).^33–34 Participation in sport among girls may provide a unique experience that can help provide a set of experiences that enhances positive development during this critical stage of adolescence and lessen risky behavior.^46,49 Problematically, girls participation rates in sport are typically lower than boys³⁵ and signals that a large segment of girls may not have access to this important resource. While it is hard to determine why this gap exists, it is important to detect potential environmental factors at the school-level that contribute to this gap. Accordingly, the purpose of this study has two specific objectives. First, using school-level data from the Office of Civil Rights Data Collection (CRDC), we will examine the provision of athletic opportunities for girls between the 2009–10 and 2017–18 school years; second, we will use the school-level data from the CRDC and the Common Core of Data (CCD) to assess how several important school-level factors are associated with the provision of athletic opportunities for girls.

METHODS

Participants

The data for this study come from two major sources that collected school-level information for the 2009–10, 2011–12, 2013–14, 2015–16, and 2017–18 school years. The first source of data is the CRDC. ³⁶ The CRDC is a mandatory survey given to all public schools and has collected information on nearly 100,000 public schools in the U.S. with the purpose to determine whether educational opportunities are being provided in an equitable manner across elementary and secondary schools (the 2017–18 data file is the most current file provided by the CRDC). The second source of data is the CCD, a nationwide survey collected annually by the National Center for Education Statistics.³⁷ The CCD gathers basic school-level information on all public elementary and secondary schools in the U.S. The two major data sources were merged for the purposes of this study.

Given that only public high schools are instructed to answer questions on interscholastic sports (per the CRDC questionnaire), the results reported here pertain to public high schools that offered grades 9 through 12. After restricting the data to these parameters, the analytic sample for this study consisted of 9,769 public high schools during the 2009–10 school year, 14,111 schools during the 2011–12 school year, 13,090 schools during the 2013–14 school year, 14,122 schools during the 2015–16 school year, and 14,244 schools during the 2017–18 school year. A total of 16,911 unique public high schools were captured in the analytic sample.

Instrumentation

Athletic Opportunities for Girls:

The key measures of interest for this analysis were captured in the CRDC data. The CRDC included information on whether schools offered interscholastic sport, the number of different sports, the number of different sport teams (e.g., J.V. football, varsity football), and individual participants as reported by school administrators (separately for girls and boys). Accordingly, the analysis included five specific measures of athletic opportunities at the school-level: (1) the school indicated not offering interscholastic sports, (2) the total number of sports schools provided to girls, (3) the total number of sport teams schools provide to girls, (4) schools’ sport participation rate for girls (i.e., a continuous measure that takes the total number of girls who participate in single-sex sports divided by the total number of girls within each school), and (5) schools sport participation gender equity ratio (i.e., a continuous measure that takes the school’s girls sport participation rate divided by the school’s boys sport participation rate).

School-Level Covariates:

Several indicators were used to assess differences based on key school characteristics. The CCD provides measures of several school-level characteristics of high schools that were used for the current study.^34,38 These school-level characteristics included racial composition, federal lunch enrollment as a proxy for school-level poverty,³⁹ sex composition, type of community where the school is located, number of students, whether the school was identified as a charter or magnet school, geographic region, and the year the school-level data was collected. Please refer to Table 1 for more descriptive information on these covariates.

Table 1:

School-Level Characteristics among Public High Schools.

	2009–10			2011–12			2013–14			2015–16
School-Level Characteristics	n = 97691			n = 141111			n = 130901			n = 141221
	%	SE		%	SE		%	SE		%	SE
Percent Female
0% to 44%	13.1%	.003		14.5%	.003		15.3%	.003		16.1%	.003
45% to 54%	81.5%	.004		78.9%	.003		78.5%	.004		77.0%	.004
55% to 100%	5.4%	.002		6.6%	.002		6.2%	.002		6.9%	.002
Percent Non-White
0% to 24%	39.6%	.005		42.9%	.004		38.4%	.004		38.6%	.004
25% to 49%	21.4%	.004		21.0%	.003		22.4%	.004		21.8%	.003
50% to 74%	15.5%	.004		14.5%	.003		15.6%	.003		15.2%	.003
75% to 100%	23.5%	.004		21.6%	.003		23.6%	.004		24.4%	.004
Percent Free Lunch Eligible
0% to 24%	37.9%	.005		36.2%	.004		28.4%	.004		28.0%	.004
25% to 49%	36.6%	.005		37.5%	.004		38.7%	.004		37.4%	.004
50% to 74%	19.2%	.004		18.7%	.003		22.9%	.004		22.2%	.004
75% to 100%	6.3%	.002		7.6%	.002		10.0%	.003		12.3%	.003
Type of Community
City	27.2%	.005		21.5%	.003		22.9%	.004		23.1%	.004
Suburb	29.3%	.005		24.9%	.004		29.7%	.004		29.2%	.004
Town	14.5%	.004		15.7%	.003		16.9%	.003		16.7%	.003
Rural	29.0%	.005		37.8%	.004		30.6%	.004		31.0%	.004
Number of Students 2
0 to 205	17.0%	.004		23.1%	.004		21.7%	.004		23.2%	.004
206 to 430	12.1%	.003		17.1%	.003		16.7%	.003		17.3%	.003
431 to 900	17.3%	.004		21.5%	.003		21.8%	.004		21.4%	.003
901 or more	53.6%	.005		38.3%	.004		39.9%	.004		38.2%	.004
Charter/Magnet School
Regular public high school	93.2%	.003		89.7%	.003		89.8%	.003		93.3%	.002
Charter/Magnet School	6.8%	.003		10.3%	.003		10.2%	.003		6.7%	.002
Special Educational School 3
Regular public high school	86.9%	.003		86.3%	.003		87.7%	.003		87.4%	.003
Special Educational School	13.1%	.003		13.7%	.003		12.3%	.003		12.6%	.003
Geographic Region
East	13.4%	.003		14.1%	.003		13.7%	.003		14.5%	.003
Midwest	23.0%	.004		28.2%	.004		23.1%	.004		25.7%	.004
South	38.0%	.005		32.9%	.004		36.7%	.004		35.1%	.004
West	25.6%	.004		24.7%	.004		26.5%	.004		24.7%	.004

% = percent; SE = Standard Error; n = sample size of schools for the specified year.

¹Since the majority of the schools used for this analysis participated across several of the school years (44.2% of schools participated in all four school years, 27.8% of the schools participated in three of the school years, and 13.9% participated in two of the school years)

²We use enrollment cutoffs that are the classifications used by the Michigan High School Athletic Association to determine which division schools will be placed (class A is the largest and class D is the smallest) (For more information: https://www.mhsaa.com/).

³Defined by the CCD as regular school versus ‘special education school/vocational school/alternative school/reportable program

Procedure

The Office for Civil Rights (OCR) conducts this survey with all U.S. public schools. It is estimated to take 14.1 hours to complete for each school. The CRDC is a mandatory data collection, authorized under the statutes and regulations implementing Title VI of the Civil Rights Act of 1964

Data Analysis

The analysis is divided into three specific sections. First, descriptive statistics for each year were provided for athletic opportunities for girls and boys. Second, we fitted both logistic and linear regression models using the generalized estimating equations (GEE) methodology with an exchangeable correlation structure to assess the association between athletic opportunities for girls and key school-level characteristics.⁴⁰ It should be noted that GEE was the optimal approach in order to retain the full population of public high schools and account for schools that participated multiple years. Third, using the same school-level characteristics as covariates in the GEE models, we then assessed how these different types of athletic opportunities provided to girls and boys (e.g., number of sports) were associated with each athletic opportunity outcome in order to assess if adding sports or sport teams for either girls or boys increased/decreased sport participation rates for girls.

RESULTS

Table 2 shows roughly 22% of schools indicated not offering any interscholastic sport. The percentage of schools not offering interscholastic sport between the 2009–10 (22.0%) and 2017–18 (19.9%) significantly declined during this eight year period (linear trend: AOR=.976, p<.001). When assessing the sample of schools that provided interscholastic sports, we see that the average number of sports offered to both girls and boys is roughly 7 to 8 sports; the average number of sport teams offered to both boys and girls is roughly 12 to 13 teams. It should be noted that small declines across the study period were found in the number of sports (linear trend: b= −.051, p<.001) and sport teams offered to boys (linear trend: b= −.259, p<.001), however, a small increase in the number of sports offered to girls was found during this eight year period (linear trend: b= .106, p<.001). The average participation rate in sport for girls was between 40.0% and 46.1%, while the average participation rate in sport for boys was between 50.9% and 57.8%. Overall, sport participation rates for girls slightly increased between the 2009–10 and 2017–18 school years (linear trend: b= .002, p<.001), while the average participation rate in sport for boys slightly decreased between the 2009–10 and 2017–18 school years (linear trend: b= −.010, p<.001). Finally, the sport participation gender equity ratio was between .759 and .871, indicating that girls sport participation rates were roughly 24% to 13% lower than boys sports participation rates across schools; the sport participation gender equity ratio did increase between 2009–10 and 2017–18 school years (linear trend: b= .025, p<.001).

Table 2:

Athletic Opportunities Provided to Girls and Boys Across Public High Schools.¹

All Schools (n = 16910; unique schools)	2009–10		2011–12		2013–14		2015–16
	n = 9769		n = 14111		n = 13090		n = 14122
	% (SE)		% (SE)		% (SE)		% (SE)
School Indicated not Having Interscholastic Sports2	22.0% (.004)		21.2% (.003)		22.6% (.003)		23.5% (.004)
Schools with Interscholastic Sports (n = 13005; unique schools)	2009–10		2011–12		2013–14		2015–16
	n = 7615		n = 11105		n = 10051		n = 10804
	Mean (SE)		Mean (SE)		Mean (SE)		Mean (SE)
Athletic Opportunities for Girls 3
Number of sports (girls)	7.82 (.043)		7.50 (.036)		7.39 (.040)		7.48 (.038)
Number of teams (girls)	13.8 (.088)		12.7 (.070)		12.2 (.077)		12.3 (.074)
Participation Rate (girls)	.400 (.002)		.461 (.002)		.420 (.002)		.434 (.002)
Athletic Opportunities for Boys 3
Number of sports (boys)	8.28 (.043)		7.84 (.035)		7.45 (.039)		7.49 (.037)
Number of teams (boys)	15.1 (.091)		13.7 (.074)		12.6 (.080)		12.7 (.077)
Participation Rate (boys)	.524 (.003)		.578 (.002)		.509 (.002)		.520 (.002)
Gender Equity (girls participation rate/boys participation rate)3
Sport Participation Gender Equity Ratio	.758 (.003)		.803 (.003)		.876 (.021)		.848 (.0033)

% = percent; SE = Standard Error; n = sample size of schools for the specified year.

¹Additional bivariate analyses (binary logistic

²and linear regression

³using a continuous measure for school-year) confirmed significant linear trends at a .001 alpha level for each athletic opportunity listed in the table above. It should be noted that no statistically significant difference was found between boys sport participation rate in 2009–10 versus 2015–16. Additionally, extreme cases were removed if schools indicated offering 51 or more sports (for girls or boys), 61 or more sport teams (for girls or boys), and had a participation rate that exceeded 2.01 (for girls or boys). This resulted in 292 cases being removed (i.e., 86 schools removed for 2009–10; 107 schools removed for 2011–12; 49 schools removed for 2013–14; 50 schools removed for 2015–16).

Table 3 shows the school-level characteristics associated with whether schools indicated not offering any interscholastic sports. The results show that schools with a higher percent of female students, schools that have a higher percent of students who are non-white, schools with more students who are free lunch eligible, schools located in cities, smaller schools (i.e., between 0 to 205 students) and schools that were considered charter/magnet have higher odds of not offering interscholastic sport when compared to their respective reference groups; for instance, the odds of schools where the student body is predominately non-white (75% to 100% of the student body) were roughly 1.3 times higher (AOR = 1.34, 95% CI = 1.20,1.50) to not offer any interscholastic sports when compared to schools where the student body is predominately White. Schools in the Midwest, South, and West had higher odds of not offering any interscholastic sports when compared to schools in the East. Moreover, after adjusting for these school-level characteristics, the odds of schools not offering any interscholastic sport was slightly lower in more recent years.

Table 3:

Assessing the Association between School-Level Characteristics and if Schools Do Not Offer Any Interscholastic Sports.

School-Level Covariates	School Does Not Offer Interscholastic Sports n = 162701
	%	AOR 95%CI
% Female
0% to 44%	57.0%
45% to 54%	11.4%	.733*** (.671,.800)
55% to 100%	64.0%	1.96*** (1.73,2.22)
% Non-White
0% to 24%	14.4%
25% to 49%	20.0%	1.27*** (1.14,1.42)
50% to 74%	24.8%	1.41*** (1.24,1.60)
75% to 100%	33.2%	1.32*** (1.15,1.51)
% Free Lunch Eligible
0% to 24%	13.4%
25% to 49%	17.5%	1.04 (.951,1.14)
50% to 74%	30.4%	1.01 (.908,1.14)
75% to 100%	35.7%	.890 (.771,1.02)
Type of Community
City	32.6%
Suburb	20.0%	.829*** (.738,.930)
Town	23.2%	.521*** (.456,.596)
Rural	16.7%	.279*** (.246,.317)
Number of Students
0 to 205	62.9%
206 to 430	29.3%	.397*** (.360,.438)
431 to 900	11.5%	.140*** (.125,.157)
901 or more	3.3%	.032*** (.028,.037)
Charter/Magnet School
Regular public high school	19.7%
Charter/Magnet School	50.9%	2.94*** (2.64,3.27)
Special Ed. School
Regular public high school	12.4%
Special Educational School	90.2%	16.3*** (14.4,18.5)
Geographic Region
East	11.7%
Midwest	21.2%	1.73*** (1.50,2.00)
South	18.9%	1.95*** (1.68,2.25)
West	34.2%	1.94*** (1.68,2.25)
School Year
2009–10	22.0%
2011–12	21.2%	.529*** (.482,.582)
2013–14	22.7%	.790*** (.722,.865)
2015–16	23.5%	.815*** (.745,.892)

^*p<.05,

^**p<.01,

^***p<.001; % = Percent; AOR = adjusted odds ratio; 95% CI = 95% confidence interval

¹Sample sizes vary due to missing values on either the outcome or covariates. It should be noted that the analyses that assessed if schools did not offer interscholastic sport used the full sample of schools (i.e., n = 16,910); while the analyses assessing number of sports offered to girls, number of sport teams offered to girls, girls sport participation rates, and sport participation gender equity ratios used only the sample of schools that indicated having interscholastic sport (i.e., n = 13,005). All analyses use listwise deletion with respect to missing information on both athletic opportunities and school-level characteristics.

Second, examining models 1 and 2 in Table 4 shows that the number of sports and sport teams for girls were significantly lower for schools with a high percent of students who were non-white, schools with a high percent of students who were eligible for free lunch, schools located in towns or rural areas, smaller schools (i.e., between 0 to 205 students) and schools that were considered charter/magnet; for instance, schools where 75% to 100% of students were eligible for free lunch typically offered 1 less sport for girls (b = −1.37, p<.001) and 3 less sport teams for girls (b = −2.93, p<.001) when compared to schools where 0% to 24% of students were eligible for free lunch. Schools in the Midwest, South, and West had fewer sports and sport teams for girls when compared to schools in the East (except the Midwest with respect to sport teams). It should also be noted that, after adjusting for these school-level characteristics, the number of sports and sport teams offered to girls slightly increased during this time period.

Table 4:

Assessing the Association between School-Level Characteristics and Athletic Opportunities for Girls.

School-Level Covariates	Number of Sports (Girls)1 n = 128892 Model 1			Number of Sport Teams (Girls)1 n = 128852 Model 2			Girls Sport Participation Rate1 n = 128732 Model 3			Sport Participation Gender Equity Rato1: n = 128352 Model 4–1.
	Mean	b SE	β	Mean	b SE	β	Mean	b SE	β	Mean	b SE	β
% Female
0% to 44%	5.35	Reference		7.95	Reference		0.532	Reference		1.000	Reference
45% to 54%	7.80	.219***(.062)	0.016	13.3	.514***(.109)	0.028	0.423	−.021***(.004)	−0.027	0.811	−.145***(.022)	−0.040
55% to 100%	5.45	.033 (.117)	0.001	7.59	−.165 (.209)	−0.005	0.410	−.073***(.007)	−0.055	0.754	−.257***(.040)	−0.037
% Non-White
0% to 24%	7.36	Reference		12.5	Reference		0.542	Reference		0.860	Reference
25% to 49%	8.00	.075 (.056)	0.015	13.7	.352***(.105)	0.033	0.403	−.030***(.003)	−0.041	0.819	−.024 (.016)	−0.008
50% to 74%	7.97	.212** (.071)	0.026	13.4	.460***(.134)	0.038	0.342	−.049***(.004)	−0.053	0.789	−.053* (.021)	−0.016
75% to 100%	7.06	−.308***(.080)	−0.009	11.4	−.624***(.152)	0.007	0.285	−.087***(.005)	−0.102	0.781	−.066** (.024)	−0.022
% Free Lunch Eligible
0% to 24%	8.66	Reference		15.7	Reference		0.525	Reference		0.972	Reference
25% to 49%	7.21	−.623***(.046)	−0.101	11.7	−1.73***(.084)	−0.158	0.425	−.057***(.002)	−0.14	0.866	.018 (.014)	0.008
50% to 74%	6.68	−1.04***(.066)	−0.138	10.4	−2.63***(.120)	−0.192	0.330	−.087***(.004)	−0.169	0.812	−.005 (.021)	−0.002
75% to 100%	5.76	−1.36***(.091)	−0.124	8.53	−3.00***(.165)	−0.159	0.314	−.096***(.005)	−0.134	0.792	−.053 (.030)	−0.012
Type of Community
City	8.18	Reference		13.8	Reference		0.294	Reference		0.810	Reference
Suburb	9.24	.099 (.064)	0.004	16.3	.498***(.123)	0.023	0.401	.026***(.004)	0.032	0.816	−.028 (.018)	−0.011
Town	6.84	−.586***(.076)	−0.059	11.4	−.710***(.146)	−0.034	0.466	.050***(.004)	0.066	0.799	−.062** (.022)	−0.020
Rural	6.03	−.661***(.070)	−0.089	9.64	−.888***(.134)	−0.067	0.521	.051***(.004)	0.075	0.853	−.041* (.020)	−0.017
Number of Students
0 to 205	3.62	Reference		5.03	Reference		0.673	Reference		0.972	Reference
206 to 430	5.25	1.32***(.079)	0.124	7.99	2.53***(.150)	0.123	0.545	−.125***(.005)	−0.184	0.866	−.095***(.024)	−0.030
431 to 900	6.93	2.77***(.079)	0.307	10.8	5.32***(.152)	0.291	0.457	−.205***(.005)	−0.361	0.812	−.138***(.024)	−0.052
901 or more	9.23	4.77***(.081)	0.620	16.4	10.0***(.154)	0.664	0.341	−.285***(.005)	−0.599	0.792	−.154***(.024)	−0.068
Charter/Magnet School
Regular public high school	7.56	Reference		12.8	Reference		0.440	Reference		0.823	Reference
Charter/Magnet School	6.94	−.529***(.081)	−0.02	10.1	−1.62***(.147)	−0.055	0.281	−.055***(.005)	−0.049	0.863	.076** (.027)	0.015
Special Ed. School
Regular public high school	7.57	Reference		12.8	Reference		0.434	Reference		0.825	Reference
Special Educational School	4.89	−1.76***(.160)	−0.053	7.26	−2.69***(.304)	−0.042	0.294	−.201***(.010)	−0.088	0.827	−.076 (.050)	−0.007
Geographic Region
East	9.31	Reference		15.4	Reference		0.550	Reference		0.880	Reference
Midwest	7.19	−1.00***(.070)	−0.110	13.1	−.126 (.139)	0.006	0.525	−.086***(.004)	−0.148	0.839	−.072***(.019)	−0.028
South	7.11	−1.59***(.070)	−0.195	10.9	−3.07***(.139)	−0.186	0.334	−.181***(.004)	−0.331	0.766	−.107***(.019)	−0.046
West	7.33	−1.51***(.076)	−0.153	13.2	−1.24*** (.151)	−0.062	0.401	−.115***(.004)	−0.175	0.871	−.009 (.021)	−0.003
School Year
2009–10	7.82	Reference		13.8	Reference		0.400	Reference		0.758	Reference
2011–12	7.50	.416*** (.041)	0.046	12.7	.366*** (.070)	0.024	0.461	.021*** (.002)	0.039	0.803	.027 (.017)	0.010
2013–14	7.39	.192*** (.042)	0.022	12.2	−.326*** (.071)	−0.012	0.420	.004 (.002)	0.006	0.876	.107*** (.017)	0.042
2015–16	7.48	.384*** (.042)	0.041	12.3	−.011 (.071)	−0.003	0.434	.011*** (.002)	0.019	0.848	.081*** (.017)	0.032

^*p<.05,

^**p<.01,

^***p<.001; SE = Standard Error; b = unstandardized beta coefficient; β = Standardized beta coefficient

¹Standardized beta coefficients were estimated with Ordinary Least Squares regression (GEE does not produce standardized estimates).

²Sample sizes vary due to missing values on either the outcome or covariates. All analyses use listwise deletion with respect to missing information on both athletic opportunities and school-level characteristics.

Third, model 3 in Table 4 shows that sport participation rates for girls were significantly lower for schools with a high percent of students who were female, schools with a high percent of students who were non-white, schools with a high percent of students who were eligible for free lunch, schools located in cities, larger schools (i.e., more than 206 students) and schools that were considered charter/magnet; for instance, schools where 75% to 100% of students were eligible for free lunch typically had sport participation rates for girls that were 9.5% lower (b = −.095, p<.001) when compared to schools where 0% to 24% of students were eligible for free lunch. Schools in the Midwest, South, and West had substantially lower sport participation rates for girls when compared to schools in the east. The adjusted results indicate that sport participation rates increased for girls during this time period.

Finally, model 4 in Table 4 shows that sport participation gender equity ratios were significantly lower for schools with a high percent of students who were female, schools with a high percent of students who were non-white, schools located in towns, larger schools, and schools that were considered regular schools; for instance, schools where 75% to 100% of students were non-white typically had gender equity ratios that were 2.1% lower (b = −.021, p<.01) when compared to schools where 0% to 24% of students were won-white. Schools in the Midwest and South had substantially lower sport participation gender equity ratios. Notably, sport participation gender equity ratios increased for girls during this time period.

Table 5 shows how different types of athletic opportunities provided to girls and boys (e.g., number of sports) were associated with each athletic opportunity outcome (girls outcomes) when controlling for school-level characteristics. Accordingly, the results show statistically significant positive associations across the majority of the different measures of athletic opportunities for girls. For instance, we see that for every one sport added for girls within schools, there is a corresponding 2.5% increase in the sport participation rate for girls (b = .025, p<.001). However, while the number of girls sports (b = .004, p<.05), girls sport teams (b = .004***, p<.001), and girls sport participation rates (b = .400, p<.001) are positively associated with schools sport participation gender equity ratio, the number of boys sports (b = −.008, p<.001), number of boys teams (b = −.006***, p<.001), and boys sport participation rates (b = −.359, p<.001) is negatively associated with schools sport participation gender equity ratios – this indicates that as boys get more athletic opportunities, gender equity in participation rates within schools decreases. Accordingly, while the results from this table show evidence that increasing the number of sport or sport teams within the school context is a substantial predictor of increasing participation rates for girls, gender equity in participation in sports is negatively impacted by athletic opportunities being provided to boys.

Table 5:

Assessing the Association between School-Level Opportunities to Participate in Athletics and Athletic Opportunities for Girls.

School-Level Opportunities	Number of Sports (Girls)1		Number of Sport Teams (Girls)1		Girls Sport Participation Rate1		Sport Participation Gender Equity Ratio1
Number of Sports (Girls)	Model 1: n = NA		Model 2: n = 128852		Model 3: n = 128712		Model 4: n = 128332
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	-- --	--	1.25*** (.006)	0.671	.025*** (.001)	0.404	.004* (.001)	0.015
Number of Sport Teams (Girls)	Model 5: n = 128852		Model 6: n = NA		Model 7: n = 128672		Model 8: n = 128292
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	.391*** (.002)	0.762	-- --	--	.014*** (.0001)	0.443	.004*** (.0003)	0.089
Girls Sport Participation Rate	Model 9: n = 128712		Model 10: n = 128672		Model 11: n = NA		Model 12: n = 128352
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	6.33*** (.073)	0.435	11.7*** (.130)	0.42	-- --	--	.400*** (.028)	0.091
Sport Participation Gender Equity Ratio	Model 13: n = 128332		Model 14: n = 128292		Model 15: n = 128352		Model 16: n = NA
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	−.001 (.013)	0.010	.954*** (.080)	0.057	.010*** (.001)	0.056	-- --	--
	Number of Sports (Girls)1		Number of Sport Teams (Girls)1		Girls Sport Participation Rate1		Sport Participation Gender Equity Ratio1
Number of Sports (Boys)	Model 17: n = 128882		Model 18: n = 128842		Model 19: n = 128702		Model 20: n = 128322
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	.917*** (.002)	0.921	1.16*** (.007)	0.617	.024*** (.001)	0.380	−.008*** (.001)	−0.028
Number of Sport Teams ( Boys )	Model 21: n = 128852		Model 22: n = 128842		Model 23: n = 128672		Model 24: n = 128292
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	.344*** (.002)	0.697	.877*** (.002)	0.925	.012*** (.0001)	0.395	−.006*** (.001)	−0.044
Boys Sport Participation Rate	Model 25: n = 128712		Model 26: n = 128672		Model 27: n = 128642		Model 28: n = 128352
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	4.98*** (.066)	0.363	9.48*** (.115)	0.368	.700*** (.002)	0.764	−.359*** (.025)	−0.086
Sport Participation Gender Equity Ratio	Model 29: n = NA		Model 30: n = NA		Model 31: n = NA		Model 32: n = NA
	b (SE)	β	b (SE)	β	b (SE)	β	b (SE)	β
	-- --	--	-- --	--	-- --	--	-- --	--

^*p<.05,

^**p<.01,

^***p<.001; SE = Standard Error; b = unstandardized beta coefficient; β = Standardized beta coefficient; NA/-- = not applicable due to the outcome and covariate being identical or a redundant analysis.

¹All analysis control for percent female, percent Non-White, percent eligible for free lunch, type of community, number of students, charter/magnet school status, special educational school status, geographic region, and school year. Standardized beta coefficients were estimated with Ordinary Least Squares regression (GEE does not produce standardized estimates).

²Sample sizes vary due to missing values on either the outcome or covariates. All analyses use listwise deletion with respect to missing information on both athletic opportunities and school-level characteristics.

DISCUSSION

This study highlights that roughly one out of five public high schools offering grades 9 through 12 do not offer interscholastic sports. While the results indicate that this trend in schools not offering sport has been declining over the past several years, the results do indicate that schools that cater to historically disadvantaged groups are more likely to indicate not having opportunities to participate in sport (i.e., girls and racial minorities). Moreover, this study provides further evidence that schools with lower resources and a higher proportion of racial minority students tend to have the fewest opportunities for girls to participate in interscholastic sport. In particular, the results from the analysis show that schools located in cities, schools classified as high-poverty, and schools that largely consist of racial minorities provide fewer sports for girls, fewer sport teams for girls, have lower sport participation rates for girls, and in some cases have lower sport participation gender equity ratios.

The results from the analyses also suggest that one clear way to increase sport participation rates for girls would be to add more athletic opportunities for girls. While adding a new sport would require considerable resources from the school, the results also indicated that adding an additional sport team (e.g., an additional Junior Varsity team) would have a similar impact on girls sport participation rates. Adding additional sport teams to sports already provided by the school could increase sport participation among girls without having to invest the amount of resources to add a new sport to the school’s athletic portfolio.

Finally, the results of this study also suggest that the provision of single-sex sports and sport teams for girls and boys has become more equitable in recent school years. For instance, during the 2009–10 school year, schools provided boys roughly 1 additional sport team when compared to girls (13.9 versus 15.2). However, during the 2017–18 school year, schools provided boys and girls a similar number of sport teams (12.6 versus 12.8). Despite greater equity in these athletic opportunities and the continual narrowing of the gender gap in sport participation rates, the gap still persists. Clearly, the analysis provided evidence that adding more sports and sport teams for girls was associated with increases in the sport participation gender equity ratio (i.e., narrows the gender gap). Problematically, the analysis also showed that adding more sports and sport teams for boys was associated with decreases in the sport participation gender equity ratio (i.e., widens the gender gap). While it is troublesome to suggest that girls should receive more sports than boys, potential solutions could be to add more sport teams to sports that already exist for girls within schools. Namely, redistributing the number of existing athletic opportunities that may be unevenly distributed within schools could promote greater participation among girls (e.g., there may be more opportunities for boys to play football than opportunities to play volleyball for girls).

Limitations

Despite some of the important contributions of this research, there are several limitations that should be noted. First, the current sample only includes public high schools that offer grades 9 through 12 and does not capture the wide range of schools in the U.S. (e.g., Private schools). Second, the measures for sports and sport teams only captures single-sex sports and does not capture athletic opportunities that are co-ed. Finally, sport participation rates in this study will be inflated due to double or triple counting participants across sports due to the question asking school administrators to count the total number of participants across each sport (e.g., a single student could be counted three times if they participate in three separate sports).

Conclusion

The results reported here have implications for parents and educators, as well as coaches and athletic administrators who work in sports. Most notably, many of these resources that can help improve the lives of girls are still lacking within underserved communities in the U.S. Greater effort is needed to pull more resources into schools that cater to underserved populations in order to expand their sporting curriculum so that regardless of gender or socioeconomic class, students will have access to and can benefit from sports participation.

IMPLICATIONS FOR SCHOOL HEALTH.

This study has specific implications for athletic administrators who work in sports. Namely, athletic directors or key stakeholders as it relates to interscholastic sports need to develop new and innovative strategies to provide greater opportunities for girls to participate in sports.

• Schools administrators and athletic director(s) need to consistently evaluate if sport opportunities are being equitably provided to girls and boys. This would involve a detailed audit of the number of opportunities (i.e., available slots) that each sport provides girls and boys.

• Reallocate sport opportunities to girls that may be taken away from other sports with a high capacity of participation opportunities/slots (e.g., football).

• A greater collaboration across schools to share resources/facilities between high and low resource schools; this may help facilitate greater involvement in sports among girls who attend schools in lower income communities.