Athletics,Athletic Leadership,and Academic Achievement

Abstract

This study examines the relationship between athletics, athletic leadership, and academic achievement. This is likely to be a tricky issue as athletes and athletic leaders are not likely to be a random group of students. To address this issue I control for school fixed effects and instrument the endogenous variables with height. I find that athletes perform better than nonathletes in every subject area tested by the High School and Beyond survey and that this effect appear to differ by sex and race. Based on the literature, these results are likely to be especially true for urban youths. In addition, there are large benefits from leadership on these athletic teams.

Keywords

athletics achievement sports adolescents high school

In 2002, 55% of all American male high school sophomores participated in interscholastic or varsity sports. 42.5 % of all female sophomores also participated in interscholastic or varsity sports. High school athletes come from all walks of life, with over 45% of White, African American, Asian, and American Indian or Alaskan Native students participating in varsity athletics. Over 47% of students from public, Catholic, or other private schools report participating on a varsity team (T. Snyder & Dillow, 2009).

Despite their popularity with parents, students and communities, the current difficult economic times have placed greater scrutiny on the role of athletics in public school districts. A survey of school leaders by the American Association of School Administrators found that 10% of school leaders reported having scrapped some form of extracurricular activities in the 2008 to 2009 academic year, and 28% said they had considered cuts in the next academic year. In many communities, cuts have been combined with increased fees for athletic participation (Cavanagh, 2009).

The fiscal stress on schools and school districts has placed an increasing emphasis on the question of whether high school sports serve as a compliment or substitute for academics in schools. Does participation in athletics improve student academic outcomes or do they serve to detract from the learning experience? In other words, is education a zero-sum game, where time spent on extracurricular activities as sports serves to crowd out scholastic learning at home or in the classroom?

This study examines these questions by examining the impact of participation in varsity athletics on student achievement on a wide variety of examinations of cognitive development using data from the High School and Beyond (HSB) survey. While a long literature has suggested a positive link between athletic participation and academic outcomes, much of this literature has not adequately controlled for issues of selection. Athletes are not likely to be a random cross-section of youths. They may possess qualities that mean that they would have performed well on tests even had they not participated in school sports. I tackle this issue by combining school fixed effects with instrumental variables techniques. The instrumental variable I use is height.

In addition, I use these techniques to examine the impact of athletic leadership on achievement, an area not previously explored in the literature. If education is indeed a zero-sum game, the extra burdens of leadership may be detrimental to learning. On the other hand, leadership may serve to teach students important life skills that may translate to better test scores. I find strong evidence that athletic participation improves test scores and that the effect is moderated by race and sex. There is also evidence that athletic leadership may have beneficial effects.

This article proceeds as follows. Section 2 reviews the empirical and theoretical literature on athletic participation, athletic leadership, and academic achievement. Section 3 describes the methodology used in this study. This section is followed by a description of the data I used in this study. Section 5 presents the results and section 6 concludes, discusses the results and offers suggestions for future research.

Literature Review

Theoretical Motivation

The National Federation of State High School Associations (NFHS) is unequivocal in its endorsement of high school sports. According to the NFHS, “ . . . interscholastic sports and fine arts activities promote citizenship and sportsmanship. They instill a sense of pride in community, teach lifelong lessons of teamwork and self-discipline and facilitate the physical and emotional development of our nation’s youth.” The NFHS asserts that high school activities as sports serve as compliments to the academic mission of schools. It claims that extracurricular activities teach teamwork, sportsmanship, winning and losing, and hard work. “Through participation in activity programs, students learn self-discipline, build self-confidence and develop skills to handle competitive situations,” adds the NFHS. “These are qualities the public expects schools to produce in students so that they become responsible adults and productive citizens.” This spillover effect of athletics has been labeled by Broh (2002) as the “developmental model.”

Schafer and Armer (1968) suggested that athletes may have superior academic outcomes because they receive additional academic assistance from peers, teachers, and coaches. The authors also suggested that athletic participation may enhance a student’s status with one’s peers, that higher peer status enhancing an athlete’s self-esteem (Phillips & Schafer, 1971). In addition, Rehberg (1969) has suggested that this higher peer status may offer membership into the “leading crowd,” a group of college-oriented overachievers. These peer effects have been documented empirically by Spady (1970). In schools that have “don’t pass, don’t play” policies, minimum grade point average (GPA) requirements may motivate athletes to perform better than nonathletes (Schafer & Armer, 1968).

In an influential study, Coleman (1961) proposed that athletics and academic achievement behaved as a zero-sum game. In many schools, particularly smaller schools, Coleman argued that the inherent social system meant that it was important to be an athlete or a scholar, but not both. According to Coleman (1961, p. 274), “Where the roles are separate, and a boy receives no special rewards for being an all-around boy, the star athlete will not worry about good grades, and the star student will concentrate on them. And because scholastic success is a distinct area of achievement, the boys named as best students will be the best students, not just athletes who do well in studies.” In Coleman’s model, athletics and academics act as substitutes. Athletes will specialize in athletics, to the detriment of academics, and scholars will specialize in academics, to the detriment of athletics. In this way, athletic participation would negatively impact academic achievement. Even in a school that rewarded students for being well-rounded, there is an assumption that time or energy spent on one aspect of school will crowd out another. Students become generalists because they are not able to be eminent in any particular field.¹

There is reason to believe that any differences between athletes and nonathletes are not from participation in sports but from preexisting differences between the two groups. For example, Hauser and Lupetow (1978) found that athletes had higher GPAs at the end of their high school career than at the start, but they do not gain as much as nonathletes over the years of high school. Their findings suggest that preexisting differences between the two groups explained differences in initial and final GPA. in addition, Spreitzer (1992) found that varsity athletes were more likely to come from economically privileged backgrounds, and have above-average cognitive abilities, self-esteem, grades, and test scores. Studies that do not address these preexisting differences suffer from selection bias.

The relationship between athletic leadership and academic achievement may also be driven by selection—this is true by definition—but there is also reason to believe that serving as a leader has constructive effects in academics. Acting as a leader may teach valuable teamwork and facilitation skills, skills that are brought to bear on team assignments or group study sessions. Leadership may also teach an individual how to better manage his or her time and organize activities, skills that may prove beneficial not only in the classroom but in the work force.

Athletics and Achievement

Much of the literature is in support of the NFHS’ positive view of sports participation. In a review of some of the earliest studies of this issue, Jacobsen (1931) found that 10 out of 17 studies found that the scholarship of athletes improved during participation in sports, or was better than that of nonathletes, or both. Four more found no impact of athletic participation, and only three found detrimental effects. Other correlational studies suggesting a positive link between student performance and high school athletics include Mahoney and Cairns (1997), Silliker and Quirk (1997), Zaugg (1998), Stegman (2000), Kaufmann (2002), Schley (2004), White (2005), Stanley (2006) and Sitkowski (2008). In addition, a study of public school seniors by the U.S. Department of Education suggested that those with extracurricular activities (sports has the highest levels of participation) were three times as likely to have a GPA of 3.0 or better (Mihoces, 1996). Spady, S. K. Anderson (1990) and Din (2005-2006) however, found no relationship between athletic participation and acdemic outcomes. It is important to note that most of this literature relied on simple comparison of means and diverged greatly in what they deemed an athlete.

Several studies found positive effects of sports participation in high school even after introducing controls for measures as GPA, cognitive development, and educational attainment. For example, Schafer and Armer (1968) found that even after matching on intelligence test score, social class, and type of curriculum, athletes still had higher GPAs. Other studies with control variables documenting this positive association between athletic participation and educational outcomes include E. E. Snyder and Spreitzer (1990), Marsh (1993), Fejgin (1994), McNeal (1995), Eccles and Barber (1999), Barber, Eccles, and Stone (2000), Broh and Videon (2002) and Carlson, Scott, Planty, and Thompson (2005). Even in Germany, not known as a hotbed for high school sports, participation in sports seems to have a positive effect on educational attainment (Cornelißen & Pfeifer, 2007).

The National Educational Longitudinal Survey (NELS) has been used extensively by researchers to study athletic participation. An interesting study by Eitle and Eitle (2002) was able to provide some information about the effect of participating in individual sports on educational achievement. Using data from the National Educational Longitudinal Survey, they found that students who played varsity basketball had lower cognitive tests scores and that students who played a sport besides basketball and football tended to have better grades. Football results were not significant. Another study using the NELS by Marsh and Kleitman (2003) found that participation in high school sports had salutary effects on school grades, coursework selection, homework, educational and occupational aspirations, self-esteem, university applications, subsequent college enrollment, and eventual educational attainment. In addition, varsity team sports appeared to have more benefits than intramural and individual sports. Using an individual fixed effects approach in combination with the NELS data, Lipscomb (2007) found that athletic participation is associated with a 2% increase in Math and science test scores. In addition, sports participation was associated with a 5% increase in bachelor’s degree attainment expectations.

A few studies have examined the possibility of heterogeneous treatment effects of sports. Another study using the NELS by Troutman (2007) found that females who engaged in interscholastic high school sports had higher odds of completing college than do nonathletes even after controlling for school random effects. Sabo, Melnick, and Vanfossen (1993) also found differences in postsecondary educational attainment using the HSB data along gender and racial lines. In particular, participation in high school sports was more likely to affect the postsecondary attainment of White males. Hanson and Kraus (1999) found that participation in sports increased female achievement in Math and science.

A few studies have used instrumental variables techniques to examine the relationship between athletic participation and academic achievement. D. J. Anderson (2001) used as instruments the proportion of girls and boys (separately) in the student’s school who participated in sports, who held part-time jobs, and who participated in other extracurricular activities in combination with HSB to examine the effects of high school athletics on educational outcomes and labor market earnings. She also conducted a separate analysis for the NELS. Only for White athletes were the results significant and positive.

Barron, Ewing, and Waddell (2000) used as instruments for athletic participation the size of the high school, the income of parents, the health of the student, whether the school attended was a private school, library-books-per-student, the faculty-to-student ratio, height, and weight. The authors did find some evidence that athletic participation directly affected wages and educational attainment. Stevenson (2010) used variation in the level of boys’ athletic participation across states before Title IX as an instrument for the change in girls’ athletic participation over the 1970s. With this instrument, she found that a 10 percentage point rise in state-level female sports participation resulted in a 1 percentage point increase in female college attendance.

Eide and Ronan (2001) used height—as in this study—as an instrument along with the High School and Beyond data to examine the impact of athletic participation on educational attainment. They found some evidence of differences along racial and gender lines. Athletic participation had a negative effect on the educational attainment of White male student athletes, a positive effect on the educational attainment and earnings of Black male student athletes, and a positive effect on the educational attainment of White female student athletes.

Strong research designs are important in studying this issue of athletic participation as selection is a likely problem. This study is mindful of issues of selection and builds on the Eide and Ronan article in several ways. First, I use this instrumental variable strategy to examine the impact of athletic participation on cognitive test scores. Second, I add school fixed effects, which control for differences across schools. Finally, I examine the issue of athletic leadership in addition to athletic participation, an issue relativey unexplored in the literature.

Athletic Leadership

While the literature on athletic participation and academic achievement is plentiful, the literature on athletic leadership is much more sparse by comparison. Nevertheless, the research that does exist provides some very interesting insights.

Hanson and Kraus (1999) found little impact of athletic leadership on achievement in Math and science exams in the NELS. This finding is in contrast to Carlson, et al. (2005), who also used NELS data. Their results suggested that elite athletes (team leaders and most valuable players) were more than twice as likely as nonathletes to have any postsecondary education by 2000 and to have earned a bachelor’s degree by 2000. They also had advantages over nonelite athletes.

In another study using the NELS, Rouse (2008) used three econometric approaches to identifying the impact of high school leadership (team captain or club officer) on subsequent educational attainment and postschooling wages. One of these models was an instrumental variables model. Her instruments were the proportion of a student’s peers who are leaders and an indicator variable of whether a student is the oldest child in his family interacted with an indicator of whether the student is a twin. In each model, she found a positive, statistically significant effect of high school leadership on educational attainment. A similar study by Lozano (2008) concluded that high school leadership activities predicted higher college attendance rates for all demographic groups.

Dhuey and Lipscomb (2008) examined the determinants of sports team leadership using three data sets: Project Talent (1960), the National Longitudinal Study of the High School Class of 1972, and High School and Beyond. Their main finding was that the relatively oldest students were more likely to be school leaders, including team captains. Female students were also somewhat more likely to serve as athletic leaders. Relevant to this study, height also appeared to explain leadership in varsity sports.

Athletics in Urban Schools

The literature suggests participation in sports has especially beneficial effects in urban schools. This hypothesis is mostly based on results from quantitative research. Schafer & Armer (1968) found participation in sports had the effect of increasing attendance rates and grade point averages for urban high school students. Eccles and Templeton (2002), Mahoney, Cairns, & Farmer (2003), Rehberg & Schafer (1968) discovered an association between participation in sports and college attendance. Finally, Segrave (1980, 1982) concluded urban athletes had lower rates of delinquency while Mahoney & Cairns (1997) and McNeal (1995) concluded urban athletes were less likely to drop out from school.

DeMeulenaere (2010) suggests some mechanisms behind this positive urban athletic effect. Based on observations and interviews of middle school children in an urban school district in Northern California, He found student involvement in sports promoted student success by structuring schedules, creating incentives, building confidence, developing positive adult and peer role models, and getting students to develop future aspirations. The literature in this section suggests that any positive findings from my study are likely to be even more so for urban youths.

Method

This study draws on the extensive literature on education production functions such as those presented by Hanushek (1979). Akin to a production function in manufacturing, an education production function models the process by which various student and school inputs are transformed into student outcomes (Rice & Schwartz, 2008). One of the education production functions I estimate in this study is presented in equation (1):

T E S T_{i s} = α A T H L E T E_{i s} + β S T U D E N T_{i s} + γ F A M I L Y_{i s} + δ S C H O O L_{s} + ζ_{s} + ε_{i s} .

In equation 1, a measure of achievement or test score, TEST, for student i in school s is a function of whether or not that student is a varsity athlete, ATHLETE,² a set of student characteristics, STUDENT, a set of family characteristics, FAMILY, a set of school characteristics, SCHOOL, an individual school fixed effect for school s, ζ, and an error term for student i in year s. It would have been interesting to look at the effects of different sports on academic outcomes, but the data unfortunately do not provide that information.

Simple cross-sectional estimation of, is likely to lead to bias from omitted variables. As touched upon briefly previously, athletes are not likely to be a random group of students. Their decision to participate in varsity sports may indicate they have better time management skills than the average youth. They may also be more competitive; that competitive spirit translating into superior performance on the field and in the classroom. In the program evaluation literature, this bias is called selection. In the econometrics literature, ATHLETE is considered an endogenous variable.

The “gold standard” in program evaluation involves the use of a randomized experiment. The randomized experiment is noted for its strong internal validity because it breaks the link between unobservables and the treatment. Each individual participating in the study has an equal chance of receiving the treatment (McEwan, 2008). However, a randomized experiment is very difficult to implement in this framework. It is very difficult to randomly assign participation in high school sports. Athletics requires certain basic levels of ability that are innate. One cannot assign speed or strength. A study that randomly assigned participation to some athletes and not others would also fly in the face of competition and merit. Experiments also tend to be costly and may have external validity concerns. As a result, most studies of athletic participation use quasi-experimental designs, which simulate some of the characteristics of randomized experiments but where the researcher does not control the assignment of treatment.

This study proposes two methods to mitigate issues relating to selection. The first is to control for school fixed effects. Controlling for school fixed effects means relying on intraschool variation in athletic participation to identify the effect of athletic participation on student achievement. Participation in varsity sports is not likely to be randomly distributed at either the student or school level. Some schools, perhaps more affluent schools with higher average peer ability, may also be more likely to have athletics programs or a greater share of students participating in athletics. Not controlling for this school fixed effect would bias the estimate of athletic participation through the correlation of school attendance and athletic participation.

Secondly, I make use of instrumental variables techniques to further identify the impact of athletic participation on student test scores. Instrumental variables regression allows a researcher to identify estimates of causal effects based entirely on random variation in the endogenous variable (McEwan, 2008). This though, is easier said than done.

As in Eide and Ronan, the instrumental variable used is height, measured in inches in the respondent’s sophomore year. For height to be a valid instrument for ATHLETE, it must satisfy two conditions. Height must be correlated with ATHLETE, that is:

Cov (HEIGHT, ATHLETE) \neq 0 .

This condition is fairly simply to test. The most common method of testing the strength of an instrument is through the use of a partial F-test. As discussed by Murray (2006), estimation with weak instruments is likely to lead to results even more biased than simple estimation by ordinary least squares regression. The first-stage partial F-statistic in the athletic participation model is 20.65, which indicates a strong instrument. Intuitively, with a few exceptions like gymnastics or equestrian, it is not surprising that taller athletes are likely to be better at sports and hence more likely to participate in them.

The second condition for a valid instrument is that HEIGHT must not be correlated with the error term. In other words,

Cov (HEIGHT, μ_{is}) = 0 .

Equation 3 is the trickier condition to test. With more than one instrumental variable, one can test whether or not at least some of the IVs are exogenous (Wooldridge, 2003). Clearly, with one instrumental variable, this test is not possible. Instead, I rely on intuition. Height appears to be fairly random. It seems to be determined by genetics rather than environmental causes, at least in a developed country as the United States.

I also include several interaction terms for race and gender in the regression model. These interactions allow the impact of sports participation to vary based on race or gender. In other words, they moderate the effect of participation on test scores. Because ATHLETE is endogenous, interactions with ATHLETE are also likely to be endogenous. I use the method presented by Harrison (2008), where predicted ATHLETE from the first stage is interacted with race and sex in the second stage. Her research suggests this method has the smallest bias in accounting for the potential endogeneity of interaction coefficients.

In the model of athletic leadership, an interaction term is added to equation (1) to produce equation (4):

T E S T_{i s} = θ A T H L E T E * L E A D E R_{i s} + β S T U D E N T_{i s} + γ F A M I L Y_{i s} + δ S C H O O L_{s} + ζ_{s} + ε_{i s} .

The estimate of θ represents the marginal impact of being an athletic leader versus other pupils. In equation 4, the variable of ATHLETIC*LEADER is considered endogenous and is instrumented with height in the base year. Once again, there is no weak instrument problem as the first-stage partial F-statistic on height is 36.99 in this specification. Leaders on athletic teams are also likely to be taller as their height makes them stand out from the crowd. in addition, their height probably makes them on average better athletes and they may become leaders in that way. This hypothesis is consistent with Figure 1, which suggests that there are some differences in height between nonathletes, athletes, and athletic leaders, with leaders, the tallest group. There are no interactions in this model as very few of the interactions were significant in initial analyses.

Figure 1.

Height differences among students.

All regressions are weighted using the first follow-up weight provided by HSB. I also estimate Huber-White robust standard errors adjusted for clustering by institution.

Data

The data for this study come from the High School and Beyond Survey, sophomore cohort. The HSB is published by the National Center for Education Statistics (NCES) and contains data on students, parents, schools, and performance on cognitive tests. With the exception of the height variable, the data I use are from the first follow-up in 1982, when most students were in their senior and final year of high school. The data are from the restricted version of the HSB.

HSB used a multistage, stratified, and clustered design. The first follow-up sample consisted of about 30,000 1980 sophomores; all students who had been selected for inclusion in the base year survey, whether or not they actually participated, had a chance of being included in the first follow-up survey. School response rate was about 70%, with public schools having higher response rates than private or parochial schools. Base year student response rate was 92.8% and the first follow-up student response rate was 95.3% (Kling & Modi, 2000).

Variables

Examinations

One of the best aspects of HSB is the wide range of examination data that is available. HSB tested students in five different subject areas: reading, two Mathematics exams, civics, science, and vocabulary. The reading exam consisted of short passages followed by comprehension questions and a few analysis and interpretation items. In the Mathematics exams, students were asked to determine which of two quantities was greater, whether they were equal, or whether there was insufficient data to answer the question. The science exam tested science knowledge and scientific reasoning ability. The writing exam tested writing ability and knowledge of basic grammar. The vocabulary exam used a synonym format. Finally, the civics examination tested principles of law, government, and social behavior (Zahs, Pedlow, Morrissey, Marnell, & Nichols, 1995). The test scores were standardized at mean 50 and standard deviation 10 by HSB.

While cognitive test scores are a fuzzy measure of learning, they are not without their merits. Studies have shown high achievement to be associated with large earning advantages later in life (Hanushek 2002). They also draw much scrutiny and attention from the media and parents. In addition, many state accountability systems and the federal No Child Left Behind law use test scores as their primary measure for reward and sanctions.

Explanatory Variables

In addition to the measures of athletic participation and leadership, I control for several variables that have been shown by theoretical and empirical research to impact academic outcomes. I control for sex, race, participation in a bilingual, bicultural program, participation in a special education program for educational or physical handicaps in the junior or senior year, participation in Talent Search or Upward Bound, two college preparation programs, average time spent on homework per week, yearly family income, paternal and maternal education level, year of birth, and two composites measuring socioeconomic status, and school environment.

Composites

There are two linear composites used in this study, a measure of socioeconomic status and a measure of school environment. The socioeconomic status composite was generated by HSB and composed from responses from the following items:

Father’s occupation

Father’s education

Mother’s education

Family income

Material possessions of the household.

I generated the measure of school environment from responses from the following items:

Problem with students not attending school

Problem with students cutting classes

Problem with students talking back to teachers

Problem with students not obeying instructions

Problem with fights between students

Problem with threats and attacks on teachers

This school environment composite has a high degree of reliability with an average interitem correlation of 0.373 and an alpha coefficient of 0.781. Both composites were standardized with mean zero.

Height (Instrumental Variable)

Height in inches in the 10th grade is the instrumental variable used in this study. Because of the importance of height to this study, I dropped observations that did not include data on height. In addition, youths reporting heights in excess of 80 inches (203.2 cm) or 6’8” were removed from the analysis.

Descriptive Statistics

Table 1 presents descriptive statistics for each of the variables used in this analysis. Some interesting findings emerge from the descriptives. Over a third of students in the data set participated in varsity sports, with about 40% athletes serving as athletic leaders. The average student in the sample was 66.25” in 10th grade. There is some variation in the height variable, with students ranging from 36 inches (91.44 cm) to 80 inches (203.2 cm). White students represent a majority of the sample as do women. A small minority of students participated in either Talent Search or Upward Bound. There was a significant share (11.6%) of students participating in a bilingual or bicultural program in their junior or senior year, more than 4% more than the share of students participating in some form of special education program. A plurality of youths had an average of 1 to 3 hr of homework a week, came from families earning US$30,000 to US$39,999 a year, and had a mother or father with a high school degree. Most students were between 18 and 20 years old when they were surveyed in 1982.

Table 1.

Descriptive Statistics.

Variable	Mean	Standard deviation	Minimum	Maximum
Dependent variables
Reading examination score	52.354	9.656	22.600	68.500
Math Level 1 examination score	52.550	9.410	28.790	70.310
Math Level 2 examination score	52.859	9.724	29.070	74.930
Civics examination score	52.039	9.161	18.780	66.130
Science examination score	51.748	9.088	22.180	72.060
Vocabulary examination score	52.460	9.336	22.600	68.500
Explanatory variables
Athlete	0.358	0.479	0.000	1.000
Athletic leader	0.143	0.350	0.000	1.000
Male	0.473	0.499	0.000	1.000
Race
African American	0.121	0.326	0.000	1.000
American Indian, Alaskan Native, or Other	0.018	0.134	0.000	1.000
Asian or Pacific Islander	0.032	0.177	0.000	1.000
Hispanic	0.208	0.406	0.000	1.000
White	0.621	0.485	0.000	1.000
In bilingual, bicultural program in junior or senior year	0.116	0.320	0.000	1.000
In special program for educational handicap in junior or senior year	0.040	0.196	0.000	1.000
In special program for physical handicap in junior or senior year	0.039	0.193	0.000	1.000
Participated in talent search	0.037	0.188	0.000	1.000
Participated in upward bound	0.023	0.151	0.000	1.000
Average time spent on homework per week
0 hr	0.052	0.221	0.000	1.000
<1 hr	0.120	0.325	0.000	1.000
1 to 3 hr	0.275	0.446	0.000	1.000
More than 3 hr, less than 5 hr	0.201	0.400	0.000	1.000
5 hr or more, less than 10 hr	0.217	0.412	0.000	1.000
10 hr or more, less than 15 hr	0.093	0.290	0.000	1.000
15 hr or more	0.042	0.200	0.000	1.000
Yearly family income
US$7,999 or less	0.079	0.243	0.000	1.000
US$8,000 - US$14,999	0.167	0.336	0.000	1.000
US$15,000 - US$19,999	0.155	0.326	0.000	1.000
US$20,000 - US$24,999	0.157	0.328	0.000	1.000
US$25,000 - US$29,999	0.172	0.340	0.000	1.000
US$30,000 - US$39,999	0.178	0.344	0.000	1.000
US$40,000 - US$49,999	0.092	0.261	0.000	1.000
Father’s highest level of education
Less than high school or don’t know	0.304	0.455	0.000	1.000
High school graduate	0.260	0.434	0.000	1.000
Some postsecondary education	0.216	0.407	0.000	1.000
College graduate	0.107	0.305	0.000	1.000
Graduate degree	0.112	0.312	0.000	1.000
Mother’s highest level of education
Less than high school or don’t know	0.247	0.428	0.000	1.000
High school graduate	0.368	0.479	0.000	1.000
Some postsecondary education	0.238	0.423	0.000	1.000
College graduate	0.089	0.283	0.000	1.000
Graduate degree	0.057	0.230	0.000	1.000
Year of birth date (19XX)	63.769	0.498	60.000	67.000
Socioeconomic status composite	−0.007	0.741	−2.815	2.146
School environment composite	0.000	0.691	−1.931	1.325
Instrumental variable
Height (in inches)	66.253	3.803	36.000	80.000

Note. There are 10,054 observations.

Results

Cross-Tabulations

I begin with a simple comparison of means between athletes and nonathletes. These analyses can be seen in Table 2. In line with other research, on each exam, the average score of athletes is higher than the average score of nonathletes. Each of the differences is significant at a .01 level. Table 3 compares average performance between nonathletic leaders and athletic leaders. The results are striking in their consistency. On every exam, leaders perform better than nonleaders. T-tests indicate the differences are significant at a .01 level. While these results are suggestive, they lack the rigorous controls that are used in the regression results.

Table 2.

Athlete T-Test Results.

	Nonathlete (N = 6455)	Athlete (N = 3599)	Difference	\|t\|	P-Value
Reading examination score	52.112	52.788	−0.675	3.363	.001
Math Level 1 examination score	51.818	53.863	−2.045	10.503	.000
Math Level 2 examination score	51.260	52.933	−1.673	8.299	.000
Civics examination score	51.832	52.410	−0.578	3.033	.002
Science examination score	51.284	52.580	−1.130	6.872	.000
Vocabulary examination score	52.228	52.876	−0.648	3.340	.001

Table 3.

Athletic Leadership T-Test Results.

	Nonathletic leader (N = 8206)	Leader (N = 1384)	Difference	\|t\|	P-Value
Reading examination score
Math Level 1 examination score	52.498	54.562	−2.064	7.597	.000
Math Level 2 examination score	51.795	53.458	−1.664	5.881	.000
Civics examination score	52.108	53.033	−0.925	3.496	.000
Science examination score	51.732	53.243	−1.511	5.753	.000
Vocabulary examination score	52.610	53.447	−0.837	3.124	.001

Regression Results

Athletic Participation

Table 4 presents the athletic participation regression results. In all six specifications, participation in high school varsity sports is positively associated with achievement in the subjects tested by High School and Beyond. The estimates are significant in four examinations—reading, Mathematics I, science, and vocabulary—even after instrumenting with height and controlling for school fixed effects. In reading, student athletes score about 0.919 standard deviations better than nonathletes. The advantage is 0.806, 1.340, and 1.104 standard deviations in Math I, science, and vocabulary, respectively. The significance levels range from .05 in Math I to .01 in reading, science and vocabulary. These across the board positive findings support the findings of the overwhelming majority of the literature documenting positive benefits of athletic participation.

Table 4.

Instrumental Variables Regression Results, Athlete Considered Endogenous.

Variable	I	II	III	IV	V	VI
Variable	Reading	Math I	Math II	Civics	Science	Vocabulary
Athlete	9.190***	8.056**	4.619	6.403	13.395***	11.039***
Athlete	(2.61)	(2.17)	(1.12)	(1.50)	(4.20)	(3.10)
Interaction terms
Male × athlete	−1.135	1.358	3.918*	−0.083	−3.350**	−4.833**
Male × athlete	(0.62)	(0.78)	(1.73)	(0.04)	(1.98)	(2.87)
African American × athlete	−5.493**	−9.485***	−11.389***	−6.886***	−2.529	0.727
African American × athlete	(2.40)	(4.59)	(4.64)	(2.80)	(1.10)	(0.32)
American Indian, Alaskan Native, or Other × athlete	−5.697	−9.675	−18.727***	−3.297	−8.700	−11.653*
American Indian, Alaskan Native, or Other × athlete	(0.81)	(1.57)	(3.55)	(0.52)	(1.43)	(1.81)
Asian or Pacific Islander × athlete	−2.754	−2.442	−1.197	4.453	1.484	−3.292
Asian or Pacific Islander × athlete	(0.59)	(0.64)	(0.30)	(1.07)	(0.41)	(0.76)
Hispanic × athlete	1.888	−3.376*	−8.550***	−1.557	−1.836	3.464
Hispanic × athlete	(0.90)	(1.78)	(3.74)	(0.68)	(0.82)	(1.61)
Male	−0.508	0.438	1.080	−1.348	0.574	−1.080
Male	(0.71)	(0.53)	(1.37)	(1.55)	(0.89)	(1.50)
Race
African American	−5.640***	−5.707***	−3.778***	−3.561***	−8.215***	−7.135***
African American	(7.75)	(8.28)	(4.02)	(4.20)	(12.13)	(9.66)
American Indian, Alaskan Native, or other	−2.745**	−3.675***	−2.689***	−1.877*	−3.146***	−3.105***
American Indian, Alaskan Native, or other	(2.49)	(4.12)	(3.32)	(1.75)	(3.26)	(3.22)
Asian or Pacific Islander	−2.294***	0.618	0.582	−0.939	−0.670	−3.807***
Asian or Pacific Islander	(3.92)	(1.10)	(0.79)	(1.55)	(1.14)	(6.09)
Hispanic	−5.428***	−5.556***	−3.918***	−4.346***	−5.874***	−6.115***
Hispanic	(11.36)	(11.94)	(6.84)	(7.54)	(12.20)	(13.90)
In bilingual, bicultural program in junior or senior year	1.881***	1.772***	1.172***	1.739***	1.287***	2.400***
In bilingual, bicultural program in junior or senior year	(4.05)	(4.89)	(2.99)	(4.00)	(3.66)	(6.12)
In special program for educational handicap in junior or senior year	−2.857***	−1.561*	−1.576*	−2.225*	−2.477**	−3.076***
	(3.08)	(1.81)	(1.69)	(1.85)	(2.54)	(3.95)
In special program for physical handicap in junior or senior year	−0.030	0.043	0.092	−0.512	−0.281	−0.246
	(0.03)	(0.05)	(0.10)	(0.49)	(0.28)	(0.27)
Participated in talent search	0.815	0.598	0.827	0.990	−0.667	0.147
Participated in talent search	(1.15)	(0.99)	(1.21)	(1.53)	(0.89)	(0.22)
Participated in upward bound	−0.715	−0.281	0.539	−0.292	−0.366	−0.254
Participated in upward bound	(0.71)	(0.33)	(0.58)	(0.36)	(0.42)	(0.29)
Average time spent on homework per week
Less than 1 hr	1.236**	0.164	1.107	1.252*	0.181	1.279**
Less than 1 hr	(1.98)	(0.34)	(1.56)	(1.73)	(0.38)	(2.54)
1 to 3 hr	2.084***	0.897*	0.926*	1.757**	0.445	1.420***
1 to 3 hr	(3.70)	(1.81)	(1.78)	(2.37)	(0.96)	(2.70)
More than 3 hr, less than 5 hr	3.155***	2.127***	1.788**	2.151**	0.652	1.963***
More than 3 hr, less than 5 hr	(5.27)	(3.41)	(2.89)	(2.32)	(1.12)	(3.56)
5 hr or more, less than 10 hr	3.637***	3.188***	2.985***	2.859**	1.286**	2.534***
5 hr or more, less than 10 hr	(5.00)	(4.74)	(4.12)	(2.37)	(2.06)	(3.92)
10 hr or more, less than 15 hr	4.739***	4.610***	4.466***	3.309***	1.916***	3.356***
10 hr or more, less than 15 hr	(6.11)	(6.90)	(6.15)	(2.91)	(2.89)	(5.13)
15 hr or more	6.536***	6.774***	6.089***	4.988***	4.829***	5.483***
15 hr or more	(7.07)	(8.77)	(7.29)	(4.90)	(7.16)	(6.52)
Yearly family income
US$8,000 - US$14,999	1.376***	1.337***	1.001*	2.164***	2.047***	1.233**
US$8,000 - US$14,999	(2.69)	(2.66)	(1.84)	(3.83)	(4.10)	(2.22)
US$15,000 - US$19,999	1.679***	1.180**	0.961*	1.858***	1.278**	0.800*
US$15,000 - US$19,999	(2.98)	(2.29)	(1.78)	(3.29)	(2.31)	(1.68)
US$20,000 - US$24,999	1.288**	1.495***	1.381**	1.901***	2.025***	1.497**
US$20,000 - US$24,999	(2.45)	(2.88)	(2.36)	(3.01)	(3.87)	(2.89)
US$25,000 - US$29,999	1.355**	1.153**	1.076*	1.721***	1.279**	0.784
US$25,000 - US$29,999	(2.48)	(2.28)	(1.92)	(2.95)	(2.36)	(1.49)
US$30,000 - US$39,999	1.226**	0.975*	0.976*	2.138***	1.325**	0.900*
US$30,000 - US$39,999	(2.14)	(1.69)	(1.78)	(3.66)	(2.42)	(1.73)
US$40,000 - US$49,999	0.617	0.790	0.895	1.487**	0.717	0.146
US$40,000 - US$49,999	(0.80)	(1.31)	(1.42)	(2.14)	(1.15)	(0.25)
Father’s highest level of education
High school graduate	0.368	0.477	0.515	0.167	0.162	−0.077
High school graduate	(0.81)	(1.07)	(1.30)	(0.41)	(0.45)	(0.19)
Some postsecondary education	0.869**	0.831*	0.670	0.484	0.475	0.378
Some postsecondary education	(2.00)	(1.67)	(1.44)	(0.99)	(1.15)	(0.80)
College graduate	1.812***	1.778**	1.924***	1.336**	1.297**	1.342**
College graduate	(3.26)	(3.01)	(3.40)	(2.37)	(2.37)	(2.42)
Graduate degree	1.715**	0.573	1.306*	1.284*	0.163	1.147
Graduate degree	(2.29)	(0.80)	(1.69)	(1.75)	(0.22)	(1.64)
Mother’s highest level of education
High school graduate	0.568*	0.607**	0.185	0.202	0.392	0.771
High school graduate	(1.76)	(1.99)	(0.57)	(0.54)	(1.26)	(0.19)
Some postsecondary education	1.213***	0.909**	0.587	0.867*	0.859**	0.757*
Some postsecondary education	(2.83)	(2.22)	(1.06)	(1.66)	(2.03)	(1.74)
College graduate	1.027*	0.962	1.181	0.792	0.545	0.841
College graduate	(1.67)	(1.57)	(1.53)	(1.03)	(0.88)	(1.43)
Graduate degree	1.356**	0.711	0.777	1.013	1.109*	1.323**
Graduate degree	(2.08)	(1.07)	(1.00)	(1.43)	(1.77)	(2.27)
Year of birthdate (19XX)	2.169***	2.400***	1.668***	1.838***	1.839***	2.118***
Year of birthdate (19XX)	(8.79)	(10.78)	(6.11)	(7.75)	(7.82)	(6.21)
Socioeconomic status composite	0.681**	1.284***	0.915***	0.222	0.626**	1.155***
Socioeconomic status composite	(2.33)	(4.83)	(3.08)	(0.72)	(2.27)	(4.17)
School environment composite	1.427***	1.041***	1.363***	0.972***	1.176***	0.576***
School environment composite	(7.01)	(3.39)	(7.24)	(3.39)	(6.36)	(2.78)
R²	0.355	0.424	0.331	0.300	0.395	0.432

Note. There are 10,054 observations. Regressions include school fixed effects and are weighted with High School and Beyond provided weight. Robust standard errors adjusted for clustering by school. First-stage partial F-statistic on height is 20.65. *p < .10. **p < .05. ***p < .01.

Several of the interaction terms are also significant. Participation in athletics appears to benefit males more than females in Math II. However, the results are switched when it comes to the science and vocabulary examinations. On these exams, male athletes perform 0.335 and 0.483 standard deviations worse than female athletes. Race also appears to have a moderating impact on athletic participation. Athletic participation appears to be a detriment to African American youths. In Math I and II, and civics, the negative conditional effect is large enough to more than cancel out the positive main effect of athletic participation. Participation in sports appears to have a negative effect for American Indians, Alaskan Natives, or others on Math II and vocabulary. Hispanic athletes also score worse than White athletes on both Math examinations.

The other explanatory variables for the most part have the expected sign. There is strong evidence of an achievement gap between African American, American Indian, Alaskan Native, and Hispanic youths and White youths. African American students for example, score more than a standard deviation below White students in Math II. Also a disadvantage is participation in a special program for educational handicaps in the junior or senior year. There however is little to no impact of participation in a special program for physical handicaps, which makes sense as these programs are not meant to treat cognitive deficiencies in the way programs for educational handicaps do.

On every exam, there is a positive relationship between time spent on homework and performance on the HSB exams. Students who spend more than 15 hr a week on homework score on average 4.829 to 6.774 points better than students who spend zero hr a week on homework. There are general advantages for students who come from wealthy and educated families. Finally, year of birth, socioeconomic status, and school environment are all positively related to student achievement.

Somewhat surprising is the positive coefficient on bilingual or bicultural education. In every exam, including reading and vocabulary, students in a bilingual or bicultural education program junior or senior year perform better than students who do not participate in such a program. This result is significant on every test. Also surprisingly is the lack of significance in any specification of Upward Bound and Talent Search, two college preparation programs funded by the U.S. Department of Education.

The R-squared coefficients on these regression models are high and range from 0.300 in civics to 0.424 in Math I. In Math I, this value means that the regression model explains 42.4% of the variation in Math test score.

Athletic Leadership

As most of the nonathlete control variables behave in the same way as in the athletic participation models, I focus on the athletics variables in this discussion. The athletic leadership results are presented in Table 5.

Table 5.

Instrumental Variables Regression Results, Athletic Leader Considered Endogenous.

	I	II	III	IV	V	VI
Variable	Reading	Math I	Math II	Civics	Science	Vocabulary
Athletic leader	9.295**	8.365**	5.241	6.266	13.663***	11.549***
Athletic leader	(2.34)	(2.02)	(1.22)	(1.37)	(3.48)	(2.67)
Male	0.371	1.281**	1.649***	−0.683	1.905***	−0.202
Male	(0.82)	(2.35)	(3.59)	(1.31)	(4.36)	(0.42)
Race
African American	−5.398***	−5.700***	−4.029***	−3.543***	−7.359***	−6.507***
African American	(9.43)	(11.35)	(6.28)	(6.02)	(12.82)	(11.16)
American Indian, Alaskan Native, or other	−2.179**	−3.022***	−1.913**	−1.492	−2.310***	−2.188**
American Indian, Alaskan Native, or other	(2.05)	(3.55)	(2.31)	(1.35)	(2.58)	(2.16)
Asian or Pacific Islander	−2.384***	0.497	0.453	−1.196*	−0.951	−3.904***
Asian or Pacific Islander	(3.91)	(0.85)	(0.63)	(1.83)	(1.40)	(5.80)
Hispanic	−4.800***	−5.000***	−3.593***	−3.907***	−5.032***	−5.601***
Hispanic	(13.14)	(13.60)	(8.47)	(8.93)	(10.81)	(13.89)
In bilingual, bicultural program in junior or senior year	2.328***	2.187***	1.444***	2.065***	1.939***	2.929***
In bilingual, bicultural program in junior or senior year	(4.98)	(6.09)	(3.90)	(5.45)	(4.94)	(7.37)
In special program for educational handicap in junior or senior year	−2.286**	−1.088	−1.340	−1.855	−1.697	−2.384***
	(2.38)	(1.18)	(1.33)	(1.45)	(1.49)	(2.75)
In special program for physical handicap in junior or senior year	−0.455	−0.324	−0.120	−0.800	−0.834	−0.762
	(0.50)	(0.33)	(0.12)	(0.75)	(0.70)	(0.75)
Participated in Talent Search	0.989	0.780	0.961	1.123	−0.425	0.356
Participated in Talent Search	(1.34)	(1.12)	(1.36)	(1.61)	(0.41)	(0.42)
Participated in Upward Bound	−0.581	−0.182	0.580	−0.222	−0.187	−0.206
Participated in Upward Bound	(0.57)	(0.21)	(0.61)	(0.27)	(0.18)	(0.21)
Average time spent on homework per week
Less than 1 hr	1.259**	0.219	1.166*	1.291*	0.159	1.183**
Less than 1 hr	(2.11)	(0.46)	(1.73)	(1.72)	(0.30)	(2.30)
1 to 3 hr	2.336***	1.201***	1.193**	1.965***	0.722	1.706***
1 to 3 hr	(4.33)	(2.61)	(2.45)	(2.83)	(1.47)	(3.24)
More than 3 hr, less than 5 hr	3.677***	2.689***	2.239***	2.550***	1.346**	2.637***
More than 3 hr, less than 5 hr	(7.05)	(5.21)	(4.14)	(3.23)	(2.39)	(5.14)
5 hr or more, less than 10 hr	4.265***	3.852***	3.497***	3.356***	2.157***	3.379***
5 hr or more, less than 10 hr	(7.13)	(7.25)	(5.88)	(3.24)	(3.75)	(5.97)
10 hr or more, less than 15 hr	5.028***	4.949***	4.756***	3.558***	2.307***	3.83***
10 hr or more, less than 15 hr	(6.86)	(7.92)	(7.17)	(3.33)	(3.28)	(5.81)
15 hrs or more	6.629***	6.907***	6.219***	5.074***	4.873***	5.755***
15 hrs or more	(7.15)	(8.96)	(7.54)	(4.96)	(6.23)	(6.32)
Yearly family income
US$8,000 - US$14,999	1.105*	1.043**	0.753	1.950***	1.692***	1.120**
US$8,000 - US$14,999	(1.86)	(2.11)	(1.46)	(3.58)	(2.72)	(2.02)
US$15,000 - US$19,999	1.690***	1.136**	0.863	1.836***	1.323**	0.999*
US$15,000 - US$19,999	(2.72)	(2.07)	(1.57)	(3.16)	(2.02)	(1.81)
US$20,000 - US$24,999	1.232**	1.368**	1.197**	1.825***	1.962***	1.575***
US$20,000 - US$24,999	(2.13)	(2.48)	(2.05)	(2.89)	(3.10)	(2.75)
US$25,000 - US$29,999	1.474**	1.200**	1.022*	1.778***	1.479**	1.100*
US$25,000 - US$29,999	(2.42)	(2.28)	(1.80)	(3.02)	(2.13)	(1.84)
US$30,000 - US$39,999	1.334**	1.011*	0.913	2.180***	1.481**	1.233**
US$30,000 - US$39,999	(2.14)	(1.66)	(1.61)	(3.64)	(2.25)	(2.12)
US$40,000 - US$49,999	0.582	0.725	0.799	1.451**	0.681	0.234
US$40,000 - US$49,999	(0.65)	(1.04)	(1.23)	(2.08)	(0.84)	(0.33)
Father’s highest level of education
High school graduate	0.718*	0.779**	0.674**	0.408	0.675**	0.418
High school graduate	(1.80)	(2.13)	(2.16)	(1.24)	(2.02)	(1.19)
Some postsecondary education	1.397***	1.312***	0.970***	0.857**	1.243***	1.120***
Some postsecondary education	(3.81)	(3.21)	(2.65)	(2.31)	(3.36)	(2.73)
College graduate	2.451***	2.410***	2.386***	1.818***	2.224***	2.200***
College graduate	(4.86)	(4.95)	(4.99)	(3.68)	(4.21)	(4.38)
Graduate degree	2.505***	1.423***	1.995***	1.938***	1.306*	2.115***
Graduate degree	(3.75)	(2.64)	(3.41)	(3.22)	(1.83)	(3.28)
Mother’s highest level of education
High school graduate	0.780**	0.784**	0.274	0.348	0.721**	1.064***
High school graduate	(2.23)	(2.23)	(0.97)	(1.04)	(2.02)	(3.17)
Some postsecondary education	1.481***	1.147***	0.720	1.068**	1.293***	1.185***
Some postsecondary education	(3.65)	(2.88)	(1.57)	(2.40)	(3.03)	(2.86)
College graduate	1.587***	1.550***	1.648***	1.247**	1.400**	1.510***
College graduate	(2.91)	(3.01)	(2.66)	(2.03)	(2.35)	(2.73)
Graduate degree	1.835***	1.171*	1.100	1.373**	1.843***	1.862***
Graduate degree	(2.88)	(1.91)	(1.58)	(2.20)	(2.67)	(3.00)
Year of birthdate (19XX)	2.180***	2.389***	1.635***	1.837***	1.859***	2.275***
Year of birthdate (19XX)	(8.15)	(10.27)	(5.93)	(7.39)	(6.74)	(5.90)
Socioeconomic status composite	0.565*	1.177***	0.846***	0.142	0.466	1.009***
Socioeconomic status composite	(1.72)	(4.10)	(2.75)	(1.31)	(1.42)	(3.10)
School environment composite	1.436***	1.662***	1.379***	0.983***	1.199***	0.677***
School environment composite	(7.09)	(6.61)	(7.41)	(3.38)	(6.06)	(3.12)
R²	0.067	0.150	0.115	0.063	0.063	0.011

Note. There are 10,054 observations. Regressions include school fixed effects and are weighted with High School and Beyond provided weight. Absolute value of t-statistics adjusted for clustering by school. First-stage partial F-statistic on height is 36.989. *p < .10. **p < .05. ***p < .01.

The results are equally positive when it comes to athletic leadership. Athletic leaders score better than nonleaders on every exam. The estimates are only significant though, in reading, Math I, science and vocabulary. In science and vocabulary, the advantage for athletic leaders is especially large. In science, leaders score about 13 and a half points or 1.366 standard deviations higher than nonleaders. In vocabulary, leaders score 11.549 points higher than nonleaders. The gain to athletic leadership is 0.930 and 0.837 standard deviations in reading and Math I.

Conclusion

Using an instrumental variable strategy in combination with school fixed effects, this study examines the relationship between athletics and academic achievement. In addition, I use the same strategy to study any link of athletic leadership to academic achievement. The results are very intriguing.

In line with the overwhelming preponderance of findings in the literature, participation in high school sports does appear to have beneficial effects on academic achievement, as measured by cognitive test scores. These effects are especially eminent in reading, science and vocabulary. I also find some evidence of moderating effects on the impact of athletic participation on academic achievement. African American and Hispanic athletes appear to benefit less than White athletes from athletic participation. In contrast, male athletes appear to have an advantage over female athletes in terms of academic achievement. These results suggest that while there is a general benefit to athletic participation, it is far from universal. My review of the literature on school athletics suggests urban youths can particularly benefit from participation in school sports. Given the positive results above, this conclusion is still likely to hold.

My results also suggest leadership of athletic teams has great benefits. The gains to athletic leadership are larger than the gains of athletic participation, suggesting leadership has a benefit over and above the benefit of athletic participation. While serving as a leader of an athletic team, students may learn valuable teamwork, time management, and organization skills that improve their level of achievement in the classroom.

In general, I find little evidence of the zero-sum model proposed by Coleman. There is little evidence that athletics and athletic leadership serve to compete with academics. Students can have their cake and eat it too. Indeed, athletics and academics appear to reinforce each other. The skills developed as a participant or leader on an athletic team are useful in the academic sphere as well.

Suggestions for Further Research

While this study has documented a positive link between athletic participation and academic achievement, the mechanism or mechanisms behind this relationship remain unknown. This is where qualitative research can play an important role. Athletics may improve achievement through increases in human capital, but they may also work through the way others react to athletic participation. Teachers and other students may offer more counseling or high peer status may increase self-esteem. Any of these explanations is a plausible account of why athletes do better academically.

This study relied on a single year of data. Study designs utilizing longitudinal data in combination with instrumental variable techniques to study the issue of athletic participation and leadership have the potential to be very strong designs. Longitudinal data would allow for the removal of individual fixed effects, allowing for exogenous shock of joining a team or becoming a leader to help identify the effect on achievement. In addition, the data is quite old, and the conclusions may not hold for more contemporary times. This study would benefit from more recent data.

It would also be interesting to examine the effect of individual sports on academic outcomes using the strategy laid out in this study. The commitment to certain sports—football and basketball come to mind—may be greater than in other sports or may bring more esteem to the participant. Moreover, there may be differential effects from participation in team versus individual sports. Team sports may teach team building skills that individual sports do not.

Footnotes

Acknowledgements

The author wishes to thank participants at the Association for Education Finance and Policy conference and Rachana Bhatt for their helpful comments.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Ryan Yeung is an assistant professor of Public Administration at the State University of New York College at Brockport. His research focuses on education policy and public finance issues. He has a PhD in Public Administration and Policy from Syracuse University.

References

Anderson

D. J.

(2001). “If you let me play”: The effects of participation in high school athletics on students’ educational and labor market success. doi:dx.doi.org/10.2139/ssrn.258751

Anderson

S. K.

(1990). The effect of athletic participation on the academic aspirations and achievement of African American males in a New York City high school. Journal of Negro Education, 59, 507-516.

Barber

B. L.

Eccles

J. S.

Stone

M. R.

(2000). Whatever happened to the jock, the brain, and the princess? Young adult pathways linked to adolescent activity involvement and social identity. Journal of Adolescent Research, 16, 429-455.

Barron

J. M.

Ewing

B. T.

Waddell

G. R.

(2000). The effects of high school athletic participation on education and labor market outcomes. Review of Economics and Statistics, 82, 409-421.

Broh

B. A.

(2002). Linking extracurricular programming to academic achievement: Who benefits and why? Sociology of Education, 75(1), 69-95.

Carlson

Scott

Planty

Thompson

(2005). What is the status of high school athletes 8 years after their senior year? Washington, DC: National Center for Education Statistics, Institute of Education Sciences.

Cavanagh

(2009). Sour economy places athletics in jeopardy; Programs cut or scaled back, fees raised. Education Week, 28, 6-7.

Coleman

J. S.

(1961). The adolescent society. New York, NY: The Free Press of Glencoe.

Cornelißen

Pfeifer

(2007). The impact of participation in sports on educational attainment—New evidence from Germany. Economics of Education Review, 29(1), 94-103.

10.

DeMeulenaere

(2010). Playing the game: Sports as a force for promoting improved academic performance for urban youth. Journal of Cultural Diversity, 17(4), 127-135.

11.

Dhuey

Lipscomb

(2008). What makes a leader? Relative age and high school leadership. Economics of Education Review, 27, 173-183.

12.

Din

F. S.

(2005-2006). Sports activities versus academic achievement for rural high school students. National Forum of Applied Educational Research Journal-Electronic, 19(3E), 1-11.

13.

Eccles

J. S.

Barber

B. L.

(1999). Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research, 14(10), 10-43.

14.

Eccles

J. S.

Templeton

(2002). Extracurricular and other after-school activities for youth. Review of Research in Education, 26, 113-180.

15.

Eide

E. R.

Ronan

(2001). Is participation in high school athletics an investment or a consumption good? Evidence from high school and beyond. Economics of Education Review, 20, 431-442.

16.

Eitle

T. M.

Eitle

D. J.

(2002). Race, cultural capital, and the educational effects of participation in sports. Sociology of Education, 75(2), 123-146.

17.

Fejgin

(1994). Participation in high school competitive sports: a subversion of school mission or contribution to academic goals? Sociology of Sport Journal, 11, 211-230.

18.

Hanson

S. L.

Kraus

R. S.

(1999). Women in male domains: Sport and science. Sociology of Sport Journal, 16, 92-110.

19.

Hanushek

E. A.

(1979). Conceptual and empirical issues in the estimation of educational production functions. Journal of Human Resources, 14, 351-388.

20.

Hanushek

E. A.

(2002). Publicly provided education. In Auerbach

A. J.

Feldstein

(Eds.), Handbook of public economics (vol. 4, pp. 1787-2430). Amsterdam, Netherlands: Elsevier.

21.

Harrison

T. D.

(2008). On 2SLS estimation of an interaction coefficient. Unpublished manuscript, Drexel University, Philadelphia, PA.

22.

Hauser

W. J.

Lupetow

L. B.

(1978). Participation in athletics and academic achievement: A replication and extension. The Sociological Quarterly, 19, 304-309.

23.

Jacobsen

J. M.

(1931). Athletics and scholarship in the high school. The School Review, 39, 280-287.

24.

Kaufmann

A. M.

(2002). Interscholastic sports participation as a predictor of academic success for high school students (Unpublished dissertation). University of Wisconsin-Milwaukee, Milwaukee, WI.

25.

Kling

Modi

(2000, November 13). Surveys measuring wellbeing: High school & beyond. Retrieved from http://www.nber.org/~kling/surveys/HSB.htm

26.

Lipscomb

(2007). Secondary school extracurricular involvement and academic achievement: a fixed effects approach. Economics of Education Review, 26, 463-472.

27.

Lozano

F. A.

(2008). Language, high school leadership and the postsecondary outcomes of Hispanic students. Economics of Education Review, 27, 342-353.

28.

Mahoney

J. L.

Cairns

B. D.

Farmer

T. W.

(2003). Promoting interpersonal competence and educational success through extracurricular activity participation. Journal of Educational Psychology, 95, 409-418.

29.

Mahoney

J. L.

Cairns

R. B.

(1997). Do extracurricular activities protect against early school dropout? Developmental Psychology, 33, 241-253.

30.

Marsh

H. W.

(1993). The effects of participation in sport during the last two years of high school. Sociology of Sport Journal, 10, 10-43.

31.

Marsh

H. W.

Kleitman

(2003). School athletic participation: Mostly gain with little pain. Journal of Sport & Exercise Psychology, 25, 205-228.

32.

McEwan

P. J.

(2008). Quantitative research methods in education finance and policy. In Ladd

H. F.

Fiske

E. B.

(Eds.), Handbook of research in education finance and policy (pp. 87-104). New York, NY: Routledge.

33.

McNeal

R. B.

(1995). Extracurricular activities and high school dropouts. Sociology of Education, 68(1), 62-80.

34.

Mihoces

(1996, May 16). Sports, books go hand in hand: Athletes adept at getting jump on their studies. USA Today, p. 1C.

35.

Murray

M. P.

(2006). Avoiding invalid instruments and coping with weak instruments. Journal of Economic Perspectives, 20(4), 111-132.

36.

Phillips

J. C.

Schafer

W. E.

(1971). Consequences of participation in interscholastic sports: A review and prospectus. The Pacific Socological Review, 14, 328-338.

37.

Rehberg

R. A.

(1969). Behavioral and attitudinal consequences of high school interscholastic sports. Adolescence, 4, 69-88.

38.

Rehberg

R. A.

Schafer

W. E.

(1968). Participation in interscholastic athletics and college expectations. American Journal of Sociology, 73, 732-740.

39.

Rice

J. K.

Schwartz

A. E.

(2008). Toward an understanding of productivity in education. In Ladd

H. F.

Fiske

E. B.

(Eds.), Handbook of research in education finance and policy (pp. 131-145). New York, NY: Routledge.

40.

Rouse

(2008). The impact of high school leadership on subsequent educational attainment and post-schooling wages. Unpublished manuscript, University of North Carolina, Chapel Hill, NC.

41.

Sabo

Melnick

M. J.

Vanfossen

B. E.

(1993). High school athletic participation and postsecondary educational and occupational mobility: A focus on race and gender. Sociology of Sport Journal, 10, 44-56.

42.

Schafer

W. E.

Armer

J. M.

(1968). Athletes are not inferior students. Society, 6(1), 21-26.

43.

Schley

(2004). The effects of athletic participation on academic achievement (Unpublished doctoral dissertation). University of Wisconsin-Stout, Stout, WI.

44.

Segrave

J. O.

(1980). Delinquency and athletics: Review and reformulation, Journal of Sport Psychology, 2(2), 82-89.

45.

Segrave

J. O.

(1982). An investigation into the relationship between participation in interscholastic athletics and delinquent behavior. Dissertation Abstracts International, 42, 4389-4390.

46.

Silliker

S. A.

Quirk

J. T.

(1997). The effect of extracurricular activity participation on the academic performance of male and female high school students. The School Counselor, 44, 228-293.

47.

Sitkowski

L. S.

(2008). The effects of participation in athletics on academic performance among high school sophomores and juniors (Unpublished doctoral dissertation). Liberty University, Lynchburg, VA.

48.

Snyder

E. E.

Spreitzer

(1990). High school athletic participation as related to college attendance among Black, Hispanic, and White males: A research note. Youth & Society, 21, 390-398.

49.

Snyder

Dillow

S. A.

(2009). Digest of education statistics. Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

50.

Spady

W. G.

(1970). Lament for the letterman: Effects of peer status and extracurricular activities on goals and achievement. American Journal of Sociology, 75, 680-702.

51.

Spreitzer

(1992). Does participation in interscholastic athletics affect adult development: A longitudinal analysis of an 18-24 age cohort. Paper presented at the Annual Conference of the American Sociological Association, Washington, DC.

52.

Stanley

K. L.

(2006). A comparative study of the effects of athletic and musical extracurricular activities on student academic achievement at the secondary level. Unpublished seminar paper, Missouri State University, Springfield, MO.

53.

Stegman

(2000). Athletics and academics: Are they compatible? High School Magazine, 7, 36-39.

54.

Stevenson

(2010). Beyond the classroom: Using Title IX to measure the return to high school sports. The Review of Economics and Statistics, 92(2), 284-301.

55.

Troutman

K. P.

(2007). From high school jocks to college grads: Assessing the long-term effects of high school sport participation on females’ educational attainment. Youth & Society, 38, 443-462.

56.

Videon

T. M.

(2002). Who plays and who benefits: Gender, interscholastic athletics, and academic outcomes. Sociological Perspectives, 45, 415-444.

57.

White

N. B.

(2005). The effects of athletic participation on academic achievement (Unpublished masters thesis). Marietta College, Marietta, OH.

58.

Wooldridge

J. M.

(2003). Introductory econometrics: A modern approach (2nd ed.). Mason, OH: Thomson-South-western.

59.

Zahs

Pedlow

Morrissey

Marnell

Nichols

(1995). High school and beyond fourth followup methodology report. Washington, DC: National Center for Education Statistics.

60.

Zaugg

(1998). Academic comparison of athletes and non-athletes in a rural high school. NASSP Bulletin, 82(599), 63-72.