The More Things Change the More They Stay the Same:
Race on the Cover of Sports Illustrated

Eric Primm, Pikeville College
Robert R. Preuhs, Metropolitan State College of Denver
Robert M. Regoli, University of Colorado
John D. Hewitt, Grand Valley State University

Introduction
     Which of the following fictitious headlines would more quickly grab the average American’s attention: “Iran joins the nuclear club,” “Dollar hits all time low,” “Jinx is over, Cubbies win the Series,” or “(Barry) Bonds cleared of all steroid allegations?” In our estimation, the sports-related headlines would more readily move a sizable minority and perhaps even a majority of Americans to read further; for good or ill, we are a nation of sports fanatics.
     Sports are woven into the fabric of everyday life. How many people save their “sick days” in order to recover from a late night of revelry while watching their favorite team in the big game? How many people limp into work on Mondays because their aging bodies can no longer take the “pounding” absorbed in the Saturday afternoon softball game? How many parents rearrange significant portions of their lives around their children’s practice and game schedules? Scores of people recall their moment from high school or college when they made the big hit, catch, goal, tackle, etc. and were heroes (at least for a few days). Many more remember the big hit, catch, or goal made by their favorite player. Millions of us develop deep, personal connections to “our” teams and the triumphs and failures of people we have never met and know little about often become our triumphs and failures.
     Similarly, sport as a social institution is interwoven with other prominent social institutions. Billions of dollars are spent each year on everything from tickets to live sporting events and pay-per-view packages to a seemingly infinite selection of sports memorabilia to suit nearly every taste. There are more than a few high schools and colleges where parents and students alike care more about the football or basketball teams’ records than drop-out rates or falling test scores. People’s devotion to sports is often cultivated by countless childhood hours spent with mom or dad watching their favorite team or by just playing catch in the back yard.
     Sports have also played a significant part in the ideological, political, and legal battles in the struggle for civil rights in the United States. Sport, as a social institution, has both been complicit in maintaining racial boundaries and instrumental in combating these practices. This disparate relationship sport has had with civil rights and the legacy of racism is the main focus of this work. Despite the gains that have been made in recent decades, sport remains “contested terrain:” an arena where ideas about race and its meanings, or lack thereof, are continually constructed and reconstructed (Hartmann 2003, p. 452).
     With the growing body of literature pointing to declining levels of racism (e.g., Avery & Rendall 2002; Butler 1991; Downey 2007; Farkas 2003; Farley 1984; Farley & Allen 1987; Grodsky & Pager 2001; McCall 2001; Moskos & Butler 1996; Sakamoto, Wu, & Tzeng 2000; Wilson 1981; Zeng & Xie 2004) we admit we were a bit skeptical upon reading in Joe Feagin’s Racist America (2000, p. 16) that “the United States (is) a ‘total racist society’ in which every major aspect of life is shaped to some degree by the core racist realities.” Upon further investigation into the claims of Feagin and the works of other critical race theorists (e.g., Bonilla-Silva 2003; Coates, 2008; Dovidio & Gaertner 1986; Feagin & Sikes 1994) a clearer picture and better understanding of the arguments began to emerge. According to several of these theorists, the racism that is widespread today is not of the same variety as that of even a generation or two ago. In no small part because of the political and social changes that occurred as a result of the Civil Rights Movement, the overt and egregious forms of racism that are easily recognized and almost universally denounced began to wane. The gains made by the CRM did not herald the end to racial bias; rather, a new form emerged to take the place of the old.
     The “new” racism is subtle, muted, and largely hidden. It remains concealed by a “new racetalk” (Bonilla-Silva 2003). This racetalk employs the rhetoric of color-blindness and extols the ideals of individualism and meritocracy. The new racism specifically ignores race, therefore ignoring many of the problems associated with racial bias, and the simple practice of discussing racial inequities is criticized for exacerbating the problem (Bonilla-Silva 2003; Coates 2008). In this process, racial bias is transformed into an individual problem with individual causes and solutions. In the meantime, those employing the new racetalk (and the larger society) are absolved of all wrongdoing or responsibility (Bonilla-Silva 2003). The structural and institutional forms of racism go unrecognized and unacknowledged, and racism becomes narrowly and erroneously defined as merely the overt acts of racist skinheads or Klansmen (Coates 2008). Of course today, the overt forms of racism are routinely and regularly denounced, but the covert forms—if recognized at all—are seen more like “racism lite” in comparison (Bonilla-Silva 2003, p. 3). If the claims of critical race theorists are correct and racism is a continuing and pervasive problem in modern American society, racial bias will also be pervasive in its institutions. In this paper we examine sport and the sports media in order to more closely examine these claims.
Race and Sport          
     Sport was one of the first social institutions to integrate and accept minority participants. Sports fans (and even non-fans) recognize the name Jackie Robinson if for no other reason than he was the first black man to break professional baseball’s “color barrier” in 1947. Other sports experienced similar firsts both before and after Robinson’s: Jack Johnson (boxing) in 1897, Kenny Washington (football) in 1946, Althea Gibson (tennis) and Earl Lloyd (basketball) in 1950, and Lee Elder (golf) in 1975. Beyond their contributions to their respective sports, these pioneering athletes were significant because, in the words of Patrick Miller (2004), “they offered a measure of hope” (p. 147). As the walls of segregation began to crumble, minority participation in sports skyrocketed. Additionally, and perhaps more notably, white Americans accepted and celebrated minority athletes and their achievements in ways once inconceivable (Price 1997b; Thomas 2002).
     Despite the pioneering integration efforts and other progress made, the institution of sport has had a checkered past in terms of the struggle for racial equality. Historically sports helped preserve and support the racial status quo, meaning among other things, strict racial segregation. Moreover, even as the days of segregation in sports were ending, minority athletes faced restricted access. For example, Major League Baseball operated under an informal “rule” during the early 1950s known as the “50 percent color line” (Kahn 1972). Essentially, this meant that a team could field up to, but no more than, four black players at any one time. Fielding five or more black players on a nine-man team would cross that 50 percent color line and threaten the status quo of white dominance in the sport. Owners and league officials considered this an intolerable prospect for “America’s pastime” and wanted no part if it (Kahn 1972).
     Another practice that has functioned to restrict minority athletes’ participation was “stacking:” limiting (or granting) access to certain positions based on race (see Coakley 2007; Frey & Eitzen 1991; Jibou 1988; Knee 2003; Lomax 1999; Medoff 1976; Spence 2000). In various sports, some positions require high levels of intelligence, leadership, and decision-making ability while others emphasize dominant physical skills like speed, strength, and agility. Furthermore, it was commonly believed that these qualities and skills were differentially distributed according to race. In football, for instance, the quarterback is typically a team leader and acts as the offensive “general” on the field; a quarterback needs to be smart and make sound, quick decisions. Until relatively recently—because of the requirements of the position coupled with antiquated, essentialist, racist beliefs—young black athletes were actively discouraged from learning and playing the quarterback position. These same young black athletes were guided into other positions where their supposed “natural” physical skills could best be utilized: running back or defensive end, for example. Although the practice of stacking has largely abated, its residual effects will likely be felt well into the future as opinions about athletes, their roles in the outcomes of games, and contributions to their respective sports are written and remembered.
     Just as people in the general public tend to be evaluated differently based on race, so too are athletes (Helmreich 1982). For example, González and here colleagues (2007) found that people attribute the success of athletes to different factors depending on the athletes’ race. Evans (1997) reported that black athletes were more likely to be seen as arrogant or insolent by both teammates and coaches. Gabriel and his associates (1999) found that media reports of minor behavioral transgressions differed depending on the race of an athlete. The manner in which race is presented is significant because neither sport nor the sports media operate in a vacuum, they interact with other major components of society (Majors 1998). Sports magazines are excellent resources for examining the interplay that occurs in defining and redefining our beliefs about the social world considering the power of visual images in forming our perceptions of reality (Eberhardt et al. 2006; Hawkins 1998).
Current Study
     For the purposes of this paper we chose to study the oldest and most prestigious of the sports news magazines, Sports Illustrated (SI), to see whether images portrayed on its covers, in relation to race, have changed over the years. Even observers who are not intimately familiar with SI would note that there has been a substantial increase in the number of minority athletes appearing on its covers since the first issue hit newsstands in 1954. We, however, wanted to more closely examine the rate of increase, especially in relation to the proportional increase of minority athletes participating in sports in general. Specifically, we focus on the percentage of black and white athletes playing professional football and basketball who appear on SI covers, and compare that to participation rates in each sport.1 If the rate of black athletes appearing on SI covers has not increased proportionally to their rates of participation, this may indicate a continued bias in sport and the sports media.
     Clearly, sport is a topic worthy of study by sociologists (Edwards 1973; Lüschen 1980). Our analysis examines not only shifts in minority athlete portrayals and acceptance, but also the broader issue of racial bias in modern-day society. Sport and the sports media are prominent and powerful institutions that produce, define, replicate, and reflect broader societal ideals. If critical race theorists are correct and these systems operate within a new form of racism, cloaked in the rhetoric of color-blindness, it seems possible that American society as a whole could function under a similar arrangement.
Methods
     Data were obtained from a content analysis of the covers of Sports Illustrated magazine, 1954 through 2004 obtained from Sports Illustrated 50 Years: The Anniversary Book (Fleder 2004). To achieve a more coherent analysis, the study was limited to covers that featured professional basketball and football players who were the only athletes on the cover.2 Once the data were sorted using these criteria, there was a total of 802 covers. 
The Variables
1. Race: The race of the athletes was coded into two categories: black (n = 409) and white (n = 393). This was accomplished by utilizing a visual inspection of the physical characteristics of the athletes.3 If there was still a question as to the athlete’s race after the visual inspection, a more detailed investigation into the athlete’s personal biography was conducted, such as consulting sports experts and additional published resources for a confirmation of race.
2. Sport: This variable was the sport each athlete played: football (n = 491) or basketball (n = 311).
3. Representation: Data were produced to document the changing participation rates of black and white athletes in professional basketball and football. Since the percentages of black and white athletes were virtually mirror images of one another only the percentage of black athletes in each sport was used in the analysis. Lapchick’s 2003 Racial and Gender Report Card provided data from 1989 to 2002 on the racial composition for the two sports. To estimate the participation rates of blacks before 1989 in football, we examined trading cards for each sport and coded the race of each player. These data were obtained from Topps Football Cards: A Complete Picture Collection 1956-1986 (Clary & Kirshbaum 1986). However, no similar data source exists for professional basketball. Therefore, to estimate minority participation in basketball, the number of black and white players on championship-winning teams was counted. These data were derived from photos of the championship teams in The Pro Basketball Encyclopedia (Hollander 1978).4 After these data sets were combined, there still were a few years with missing information (1979-1988 in basketball; 2000 and 2002 in football). In these cases, the missing data were estimated based on the linear trend of the available data.
4. Time: This variable represents the year of publication from 1954-2004 and is coded as a count, starting with 1 for 1954.
5. Time-Squared: Preliminary analysis suggested the variable, Time, may be non-linear in nature. This interaction variable was created to test that possibility, and as the name suggests, is Time multiplied by itself.
Analysis
     The dependent variable, Race, was regressed onto the variables Sport, Representation, Time, and Time-Squared. Because of the dichotomous nature of the dependent variable, binary logistic regression was utilized. The unstandardized regression coefficients, standard errors, Wald χ2s and significance levels for each variable in the model are shown in Table 1.
(Table 1 about here)
     Examining the overall model fit and strength, we find that the likelihood of correctly predicting the race of the athlete on the cover of SI increases from 51 percent in the naïve model, to 66 percent. This represents a proportional increase in predictive accuracy of 29 percent. Moreover, an inspection of the common measures of pseudo R-squared (i.e., Cox and Snell, Nagelkerke, McFadden)5 shows that 20, 27, and 16 percent (respectively) of the variance in predicting the race of athletes appearing on SI covers can be accounted for by the independent variables in the model.
     When the coefficients for the independent variables are studied a few items stand out. First, the variable with one of the strongest influences on predicting the race of an athlete appearing on SI covers is the sport played. There is a 1.64 unit decrease in the logged odds of a black athlete appearing the cover if he is a football rather than a basketball player (p < .001). A more intuitive interpretation would be that across the 51 years of covers examined and controlling for the racial representation in each sport, if the athlete on the cover of SI was a football player, the odds of that athlete being black are 81 percent lower than if the athlete were a basketball player.
     A somewhat unanticipated finding was that the proportion of blacks playing in each sport had no significant effect on the ability to predict the race of athletes appearing on SI covers. Note, however, that the direction of the relationship is in the opposite direction of what we would have predicted. Controlling for the other variables in the equation, for each percentage point increase of black athletes in football and basketball leagues, there was a .025 decrease in the logged odds of blacks appearing on the cover of SI (p = .087). More intuitively, the odds of a black athlete appearing on SI covers are 2.5 percent lower for each percentage point increase in black representation in the sports studied. Though we are wary about attributing too much importance to this variable considering the lack of statistical significance, we believe this could be an interesting area to explore in future research: first, the variable is approaching significance and second, the coefficients are in, what “common sense” would say, the “wrong direction.”
     The last variables in the model are: Time, representing the years of publication, and the interaction term, Time-Squared. An examination of Table 1 shows that both variables achieve statistical significance (p < .001). In addition, our initial suspicions were confirmed concerning the non-linearity the Time variable. Over time there was an increased likelihood of black athletes appearing on SI covers, however, the magnitude of that increase steadily diminished until the inflection point was reached in about 1993. Thereafter, there were ever decreasing odds of black athletes appearing on the covers.
(Figure 1 about here)
     To illustrate these effects, Figure 1 presents the probabilities of a black athlete being on the cover of SI by converting the unstandardized coefficients reported in Table 1.6 The probabilities are estimated by varying the years in the 51-year time frame, while holding all other variables at their means. Figure 1 also presents the probabilities for football and basketball players separately to more clearly highlight the findings discussed above. 
     The probabilities over a few major time points in the sampling frame demonstrates the curvilinear nature of the relationship found between the probability of a black athlete being featured on the cover and time. In 1954, the probability of a black athlete, either a football or basketball player, being featured on the cover was .006, or 6 out of 1000 covers. In 1964, the probability increased to .08, or about 8 out of 100 covers. Fifteen years later, in 1979, that rose to .53, and thus SI covers were about as likely to feature black football or basketball players as white players. In about 1993, the likelihood of a black athlete being the featured football or basketball player peaks with a probability of .694, or just shy of 7 out of 10 times. However, a decade later in 2004, the probability of a black athlete being featured on the cover instead of a white athlete had dropped to .569, or about 5.7 out of 10 football or basketball covers, holding all other factors constant. This is about the same probability of a black being featured on the cover as we estimated for 1981 (.574).
     Figure 1 also highlights the findings regarding the difference in black athletes being featured on the cover across sports, and the extent to which this difference depends on time. In the early decades of our time frame, the difference attributed to the sport played is relatively small. By about 1975, however, the probability of a black athlete on the cover is .649 if they played basketball, but only .264 if they played football—a difference of .385. The difference diminishes as time goes on, until about 1993, when the gap increases once again. Thus, time not only affects the probability of a black athlete being featured on the cover, it also affects the influence of the sport played on the dependent variable. Both relationships, as we have shown, are non-linear.7
Discussion
     Our analysis provided both expected and unexpected results. One outcome that was somewhat surprising was that when controlling for representation levels and the effects of time, the sport an athlete played had a strong influence on predicting the race of those featured on SI’s covers: if an athlete played football as opposed to basketball the cover was much less likely to feature a black athlete. However, after further reflection we feel this relationship has a rather straightforward explanation. As discussed earlier, various sports have had a history of “stacking” or placing players in specific positions based on beliefs rooted in biological essentialism (i.e., whites are smarter or better leaders whereas blacks are faster or more powerful).8 The one position that perhaps was most resistant to changes in our collective beliefs about athletic and intellectual capabilities based on race, and therefore the one where the lingering effects of stacking are still readily apparent, is the position of quarterback in football. This also happens to be the position that typically receives the most attention or has the highest profile. Consequently, we feel that at least part of the difference between the sports of football and basketball in predicting the race of athletes on SI covers can be explained by the high status and profile that quarterbacks occupy coupled with the history of excluding black athletes from that position.
     Looking over our analysis, the finding that perhaps raised the most eyebrows was that racial representation in the sports examined had little influence on predicting the race of athletes featured on the covers of SI. In fact, what little effect the variable does have is in the opposite direction of what we predicted: as the level of participation or representation by black athletes in football and basketball increased, the likelihood of black athletes appearing on SI covers decreased. As noted above, this “relationship” is not statistically significant so we feel any explanations would be premature at this time; however, since the variable is approaching significance and because the nature of the relationship is counterintuitive it deserves more careful consideration and suggests an intriguing avenue for future study.
     The final overall findings involved the effects of time. In the first few years of its publication, the vast majority (90 percent) of athletes gracing SI covers were white. This is not all that surprising when we consider that: (1) the overwhelming majority of athletes in the major sports (baseball, football, and basketball) were white (81, 86, and 74 percent respectively), and (2) American society was divided sharply along racial lines (Dubois & Regoli 2006). Slowly but surely black representation levels in the major sports began to rise, and outpacing those rising representation levels was the proportion of black athletes featured individually on the front of the nation’s premiere sporting news magazine. As noted above, however, this increase as a function of time is not linear and there are a few significant points in this timeline that warrant more careful scrutiny.
     When football and basketball are combined, a slow increase in the probability of black athletes appearing on SI covers is observed until approximately 1961-1962 (see Figure 1). At this time there is an appreciable upswing in the slope of the probability curve which continues for another decade until approximately 1972. From 1972 until around 1978-1979, the probability curve is fairly linear at which point it begins to flatten until it reaches its inflection point in roughly 1993. Thereafter, the probability of blacks appearing on SI covers begins to decline with time. There are no doubt several explanations that could account for these “turning points” in the slope of the probability curve, so we will focus on only a few.
     One explanation for the rapid increase observed in our probability curve is that it occurred as a result of the growing momentum of, and support for, the Civil Rights Movement. In a very real sense, there were two Americas in 1954 when SI first hit newsstands around the country. Later, after many of the legal and social reforms advanced by the Civil Rights Movement were implemented, the ever increasing slope of the probability curve halts and remains steady for several years until the slope begins to slowly decline reaching the inflection point in 1993. We doubt the editors of SI would say their choice of who to place on their covers was based in any way on race of the athlete; however, it is reasonable to assume that SI’s editors were sensitive to the shifting attitudes and demands of its readers.
     Perhaps the changes observed on the covers of SI were, at least partially related to the “new” racial bias described by Bonilla-Silva (2003) and others. Similar to the 50-percent-rule described earlier, there appears to be a critical “tipping” or “saturation” point, after which there are diminishing returns in terms of black athletes appearing on the magazine’s covers. Whether this is related to the preferences of SI’s (largely white) readership, the popularity of certain athletes, or some other explanation, it seems as though a subtle bias may be at work. As a society, issues of race today tend to be discussed in terms of color-blindness ignoring race and skin tone (Bonilla-Silva 2003; Hunter 2005). Yet, the avoidance of race does nothing to address lingering problems associated with race. Examples abound of successful minority athletes; however, this success does not necessarily mean that American society has transcended race, nor is the good fortune of the few indicative of the lived experiences of the many (Rowe 1999; Spence 2000).
     Sport and the sports media have been instrumental and important arenas in the struggle for civil rights. They have also been complicit in maintaining the racial status quo. Our analysis suggests that great progress was made during SI’s first quarter-century of publication in terms of the featuring of black athletes on the cover of its magazine. During the next 26 years this trend slowed and began to reverse. According to Scott Price, the fact that "a white majority calmly accepts minority status in one of its most cherished social institutions [sport] is itself a measure of progress" (1997b, p. 33). Perhaps Price is only partially correct: the white majority may accept minority status in sports, but only to a point.
     This work indicates several areas worth further study. First, this paper examines only black and white athletes. Future work should go beyond issues of black and white. In what ways are the portrayals of Latinos, Asian, or Native American athletes similar or different than the ones we found for black and white athletes? Second, the curious non-influence of representation levels discussed above deserves further study. In addition, we limited our work to football and basketball players. Does a different pattern emerge when baseball, soccer, or other sports are added? Do gender and race intersect in decisions regarding opportunities for being placed on the cover of SI? A separate analysis of race of female athletes included on the covers, either alone or within groups, seems warranted. Another area to explore would be to expand the study beyond the covers and pages of Sports Illustrated. How is race portrayed in other sports magazines or newspapers? Is there a difference between the print media and television in the way race is presented? We feel that these and other question can offer a more thorough understanding not only of race and subtle racial bias in sports, but the manner in which these phenomena impact, and operate in, contemporary society as a whole.
Notes
1.  To make meaningful comparisons we limited this study to black and white athletes who played basketball or football. For all three sports combined (baseball, football, and basketball) there were a total of 35 athletes of Latino, Asian, Native American, or other ancestry, 27 of which were Latino. This was out of a total of 1,171 covers. Furthermore, there were only 6 athletes from these racial categories appearing individually on SI covers prior to 1979. These numbers were simply too small for any substantive analysis.
2.  Group photographs or multiple separate photographs of players on the same cover were excluded. Covers that did not portray people (e.g., dogs, birds, horses, boats, or only text) were also excluded.
3.  While this coding scheme is not a perfect measure of the racial identification the athletes may have of themselves, it does reflect the way the athletes are perceived by the general public (i.e., Tiger Woods and Derek Jeter were coded as black) and is one of the most common methods of coding race in sport research (Brown & Bear 1999; Hewitt, Muñoz, Jr., Oliver, & Regoli 2005; Primm, DuBois, & Regoli 2007; Regoli, Hewitt, Muñoz, Jr., & Regoli 2004).
4.  This method of producing data for the race of professional basketball players may reflect some bias; however, the trend of minority participation based on this method was similar to the trends we observed in football.
5.  The Cox and Snell and Nagelkerke pseudo R-squared are provided in the standard SPSS printout and the McFadden pseudo R-squared is based on a simple calculation with information provided in the same printout (see Cox & Snell 1989; McFadden 1974; Nagelkerke 1991).
6. Transformation of the unstandardized coefficients to probabilities is based on the following equation: Prob(Black on Cover) = (exp^{Constant – 1.641*Sport-football)  - .025*(Representation) + .315*(Time) -.004*(Time-Squared)})/(1+ exp^{Constant – 1.641*Sport-football)  - .025*(Representation) + .315*(Time) -.004*(Time-Squared)}). Sport-football was set to its mean (.612), 1 or 0 for all athletes, football and basketball players, respectively; Representation was set to 55.717 (its mean); the Constant is -2.987; and Time varies from 1 to 51.
7.  The sample includes observations over a number of years and thus there is a potential for serial autocorrelation to bias the results reported in Table 1. Since the number of SI covers in each year varies, and there is no consistent interval between observations, we cannot conduct a traditional time series model to control for these effects. However, we did conduct a fixed-effects analysis that included yearly dummy variables in the model to account for yearly effects and the substantive results remained the same as reported in Table 1. Additionally, we collapsed the dataset into yearly aggregate
observations for the fifty-one years in the data to examine the model using a less precise method, but one that can account for AR1 effects. Durbin-Watson and Breusch-Godfrey tests for this aggregate model suggest that there is no serial autocorrelation in the data. Moreover, the results from this aggregate analysis were substantive the same as those reported in Table 1 even when accounting for an AR1 process. The results of both of these alternative methodological approaches confirm the robustness of the results reported.
8.  Some researchers have examined the arguments surrounding the beliefs in, and origins of, biological differences between blacks and whites (see Helmreich 1982; Price 1997a; Sailes 1991; 1998), whereas others scrutinize more nuanced explanations that the athletic prowess exhibited by blacks is a result of cultural influences (see Hoberman 1997; Majors 1998; Spence 2000; St. Louis 2004).

References

Avery, R., & Rendall, M. (2002). Lifetime inheritances of three generations of whites and blacks. American Journal of Sociology, 107, 1300-1346.

Bonilla-Silva, E. (2003). Racism without racists: Color-blind racism and the persistence of racial inequality in the United States.Lanham, MA: Rowman and Littlefield.

Brown, J., & Bear, G. (1999). Minorities in major league baseball: 1952-1987. International Review for the Sociology of Sport, 34, 411-422.

Brown v. Board of Education of Topeka, 347 U.S. 483 (1954).

Butler, J. S. (1991). Entrepreneurship and self-help among black Americans. New York: State University of New York, Albany.

Clary, J., & Kirshbaum, L. (1986). Topps football cards: The complete picture collection, 1956-1986.New York: Warner Books.

Coakley, J. (2007). Sport in society (9th ed.). New York: McGraw-Hill.

Coates, R. (2008). “Covert racism in the U.S. and globally.” Sociology Compass, 2, 208-231.

Cox, D. E., & Snell. J. E. (1989). Analysis of binary data (2nd ed.). London: Chapman and Hall.

Dovidio, J., & Gaertner, S. (1986). The aversive form of racism. In J. Dovidio & S. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 61-90). Orlando: Academic Press.

Downey, L. (2007). U.S. metropolitan-area variation in environmental inequality outcomes. Urban Studies, 44, 953-977.

DuBois, S., & Regoli, R. M. (2006). Sport and Sports Illustrated: Reinforcing racial inequality or breaking racial boundaries?” Paper presented at the annual meeting of the Midwest Sociological Society, Omaha.

Eberhardt, J. L., Davies, P. G., Purdie-Vaughns, V. J., and Johnson, S.L. (2006). Looking deathworthy: Perceived stereotypicality of Black defendants predicts capital-sentencing outcomes. Psychological Science, 17, 383-386.

Edwards, H. (1973). Sociology of sport.Homewood, IL: Dorsey Press.

Evans, Jr., A. (1997). Blacks as key functionaries: A study of racial stratification in professional sport. Journal of Black Studies, 28, 43-59.

Farkas, G. (2003). Cognitive skills and non-cognitive traits and behaviors in stratification processes. Annual Review of Sociology, 29, 541-562.

Farley, R. (1984). Blacks and whites: Narrowing the gap? Cambridge: Harvard University Press.

Farley, R., & Allen, W. (1987). The color line and the quality of life in America. New York: Russell Sage Foundation.

Feagin, J. (2000). Racist America.New York: Routledge.

Feagin, J., & Sikes, M. (1994). Living with racism: The black middle class experience. Boston: Beacon Press.

Fleder, R. (Ed.). (2004). Sports Illustrated 50 years: The anniversary book.New York: Sports Illustrated Books.

Frey, J. H., & Eitzen, S. D. (1991). Sport and society. Annual Review of Sociology, 17, 503-522.

Gabriel, P. E., Johnson, C. D., & Stanton, T. J. (1999). Customer racial discrimination for baseball memorabilia. Applied Economics, 31, 1331-1335.

González, L., Jackson, E. N., & Regoli, R. M. (2007). The transmission of racist ideology in sport: Using photo-elicitation to gauge success in professional baseball. Journal of African American Studies, 10, 46-54.

Grodsky, E., & Pager, D. (2001). The structure of disadvantage: Individual and occupational determinants of the black-white wage gap. American Sociological Review, 66, 542-567.

Hartmann, D. (2003). What can we learn from sport if we take sport seriously as a racial force? Ethnic and Racial Studies, 26, 451-483.

Hawkins, B. (1998). The dominant images of black men in America: The representation of O.J. Simpson. In G. Sailes (Ed.), African Americans in sport (pp. 39-52). New Brunswick, NJ: Transaction Publishers.

Helmreich, W. (1982). The things they say behind your back.New York: Doubleday & Company, Inc.

Hewitt, J. D., Muñoz, Jr., R., Oliver, W. L., & Regoli, R. M. (2005). Race, performance, and baseball card values. Journal of Sport & Social Issues, 29, 411-425.

Hoberman, J. (1997). Darwin’s athlete: How sport has damaged black America and preserved the myth of race.Boston: Houghton Mifflin.

Hollander, Z. (1978). The pro basketball encyclopedia. Los Angeles, CA: Corwin Books.

Hunter, M. (2005). Race, gender, and the politics of skin tone. New York: Routledge.

Jibou, R. (1988). Racial inequality in a public arena: The case of professional baseball. Social Forces, 67, 524-534.

Kahn, R. (1972). The boys of summer.New York: Harper and Row.

Knee, S. (2003). Jim Crow strikes out: Branch Rickey and the struggle for integration in American baseball. Sport in Society, 6, 71-87.

Lapchick, R. (2003). 2003 racial and gender report card.Orlando, FL: Institute for Diversity and Ethics in Sport.

Lomax, M. E. (1999). The African American experience in professional football. Journal of Social History, 33, 163-178.

Lüschen, G. (1980). Sociology of sport: Development, present state, and prospects. Annual Review of Sociology, 6, 315-347.

Majors, R. (1998). Cool pose: Black masculinity and sports. In G. Sailes (Ed.), African Americans in sport (pp. 15-22). New Brunswick, NJ: Transaction Publishers.

McCall, L. (2001). Sources of racial wage inequality in metropolitan labor markets: Racial, ethnic, and gender differences. American Sociological Review, 66, 520-541.

McFadden, D. (1974). The measurement of urban travel demand. Journal of Public Economics, 3, 303-328.

Medoff, M. H. (1976). Racial segregation in baseball: The economic hypothesis versus the sociology hypothesis. Journal of Black Studies, 6, 393-400.

Miller, P. (2004). The anatomy of scientific racism: Racialist responses to black athletic achievement. In P. Miller & D. Wiggins (Eds.), Sport and the color line (pp. 327-344). New York: Routledge.

Moskos, C., & Butler, J. S. (1996). All that we can be. New York: Basic Books.

Nagelkerke, N. J. D. (1991). A note on general definition of the coefficient of determination. Biometrika, 78, 691-692.

Price, S. L. (1997a). Is it in the genes? Sports Illustrated, 53-56,December 8.

Price, S. L. (1997b). Whatever happened to the white athlete? Sports Illustrated, 30-52,December 8.

Primm, E., Dubois, S., & Regoli, R. M. (2007). An exercise in subtleties and the transmission of racism: An analysis of Sports Illustrated covers. Journal of African American Studies, 11, 239-250.

Regoli, R. M., Hewitt, J. D., Muñoz, Jr., R., & Regoli, A. M. (2004). Location, location, location: The transmission of racist ideology in baseball cards. The Negro Educational Review, 55, 75-90.

Rowe, D. (1999). Sport, culture, and the media. Philadelphia: Open University Press.

Sailes, G. (1991). The myth of black sports supremacy. Journal of Black Studies, 21, 480-487.

Sailes, G. (Ed.). (1998). African Americans in sport. New Brunswick, NJ: Transaction Publishers.

Sakamoto, A., Wu, H., & Tzeng, J. M. (2000). The declining significance of race among American men during the latter half of the twentieth Century. Demography, 37, 41-51.

Spence, C. (2000). The skin I’m in: Racism, sports, and education. New York: Zed Book Ltd.

St. Louis, B. (2004). Sport and common-sense racial science. Leisure Studies, 23, 31-46.

Thomas, R. (2002). They cleared the Lane: The NBA’s black pioneers. Lincoln, NE: University of Nebraska Press.

Wilson, W. J. (1981). The declining significance of race. Chicago: University of Chicago Press.

Zeng, Z., & Xie, Y. (2004). Asian-Americans’ earnings disadvantage reexamined: The role of place of education. American Journal of Sociology, 109, 1075-1108.

Table 1
Regression Coefficients for Predicting Race on Sports Illustrated Covers

Cox and Snell pseudo R2 = .204
Nagelkerke pseudo R2 = .272
McFadden pseudo R2 = .164
a Basketball players are reference group.