“An Examination Of A National Football League College Draft Publication: Do Racial Stereotypes Still Exist In Football?”

JR Woodward, PH.D.
Department of Sociology and Anthropology
Montana State University
2-128 Wilson Hall
Bozeman, MT 59717
woodward@montana.edu

13 October 2002

Key Words: race, sport, football, stacking

Abstract

This paper examines the stereotypes that have been associated with the abilities of football players of different races. This practice often leads to ‘stacking’, which Coakley (1998) defines as “players from a certain racial or ethnic group being either over- or under-represented at certain positions in team sports” (p. 257). Traditionally, this has meant that African Americans were less likely to be found in ‘thinking’ positions in football, most notably at quarterback, offensive center, and inside linebacker. The author attempts to discern whether these stereotypes still exist among those most intimately involved with the issue: college and professional scouts. These issues are explored by using a content analysis of National Football League (NFL) draft periodicals published by a popular sports magazine in the United States. The results of the analysis show that scouts remain significantly more likely to describe African American players in terms of their physical/athletic capabilities, and white players in terms of their mental/intelligence abilities.

I. Introduction

1.1       As one of the main loci of theoretical and empirical investigation into the sporting world, sociology has typically applied traditional concepts within the field to the sociology of sport. Concepts such as race, class, gender, political economy, and socialization have been applied to, and illustrated by, the various contexts that sport can take. Ball (1973) has called this “a sociology through sport, using sport data to address more general sociological questions” (p. 97, italics in original). One of the more fruitful and socially relevant topics looked at through sport has been that of race. Specifically, the concept of ‘racial stacking’ has occupied sport sociologists interested in race for roughly the past 35 years.

1.2       Racial stacking refers to the over or under- representation of certain racial group members in particular positions within team sports. Therefore, ascribed characteristics rather than simply achieved characteristics influence the process of assignment to positions, even though the general public typically assumes that all sport positions are conferred on the basis of achievement only. The occurrence and degree of stacking have been well documented by researchers, though there is no firm agreement on the underlying cause(s) of this phenomenon.

1.3       A possible explanation for stacking that has been quite popular over the years is to point to certain biological attributes that might be found in particular races and not others. Presumably these attributes are unique to particular races and are causal in terms of performance differentials. Although this has been the view of both some scholars and laypersons alike for quite some time, most sport sociologists reject the attempt to find explanation in biology or genetics. Indeed, such an attempt itself might actually be considered racist. As Coakley (1998) has pointed out, most physiological studies on racial differences in sport only set out to explain why African Americans1 excel in certain sports and not why whites excel in others. The implication is that since whites are not superior in a certain field, there must be a physiological reason, not a social, mental, or emotional one. Furthermore, David Hunter (1998) has provided an excellent review of the physiological research on race and sport, and concluded that genetic racial differences simply do not tell us anything about athletic performance. In his review of studies relating to nerve conduction velocity, aerobic and anaerobic power, skeletal muscle physiology, and anthropometry and morphology, he states:

“…If African American superiority in a sport such as basketball is explainable by some physical advantage possessed by African Americans more than Caucasians, two conditions would have to be obtained. First, researchers would have to discover some physical variable, or more likely combination of variables, that are associated with basketball excellence for all performers, independent of racial background. Second, this variable or combination of variables would have to be associated with race…To date, such variables have not been established (p. 96-97, italics mine).

1.4       Having rejected potential biological explanations of racial stacking, it is now possible to move on to sociological explanations. These can be grouped as follows: outcome control; self-segregation due to attractiveness of opportunities; self-selection through role modeling; and the centrality and interaction thesis. Each will be discussed briefly, starting with centrality.

1.5       The now classic study by Loy and McElvogue (1970) is a seminal contribution to research on the racial make-up of positions in baseball and American football. Using Grusky’s (1963) ideas that racial discrimination is most likely to occur at central positions in any social organization (where the most interaction occurs), they conclude that positions of ‘centrality’ are positively related to racial segregation in professional team sports. ‘Centrality’ is defined, in terms of American football, as being the positions of quarterback, center, right guard and left guard on offense, and the three linebacker positions on defense. Their findings suggest that these central positions are indeed significantly more likely to be filled by whites than by African Americans, supporting their hypothesis that the interaction potential of the playing positions in question leads to the disproportionate representation of one race over another. It should be noted however, that in a later work, Loy questioned the utility of applying Grusky’s model to professional football; rather, the model appears to be a good fit for baseball, but not necessarily for other sports. He states, “McElvogue and I were never enchanted with our forced fit of Grusky’s model to professional football, and we have found it virtually impossible to apply the model to many other kinds of sport” (Loy, 1975, p.157). Notwithstanding this admission, the centrality thesis is important to the stacking literature because it provided a theoretical and empirical attempt to make sense of the over-representation of whites in certain football positions.

1.6       Another possible explanation for racial stacking could be African American self-selection into particular positions (Eitzen & Sanford, 1975; Olsen, 1968; Coakley, 1998). African American athletes might choose a particular position based on role models in that position. For example, if an African American child decided to play football, there would be very few punters or kickers of his (her) race to emulate. Role modeling, particularly as more and more African American athletes are marketed through television, might in fact have an impact on what positions young athletes choose at the beginning of their football careers (McPherson, 1975).

1.7       A second possibility leading to the ‘self-segregation’ of African Americans into certain positions in professional football could be the weighing of chances and opportunities. The list of African American players who have started out in the quarterback position in high school or college, only to change positions (usually to wide receiver or defensive back, both ‘non-white’ positions) before entering the NFL draft is undoubtedly long. This finding has been clearly demonstrated by Eitzen and Sanford (1975). In other words, an African American quarterback at the high school level, realizing that the chances of playing quarterback at the collegiate level are slim (although improving recently), and even slimmer at the professional level, might decide of his own accord to put his talents to use in a different position; a position more consistent with prevailing stereotypes.  Additionally, this decision to switch positions might be made for the quarterback by the coach, for much the same reasons.

1.8       Finally, some researchers (Edwards, 1973; Eitzen & Sandford, 1975; Johnson & Johnson, 1995) conclude that the relative centrality of the particular position is only important to the degree that it controls the outcome of events. This position, often called the ‘outcome control hypothesis,’ was first applied to sport by Harry Edwards (1973). Edwards maintains that African Americans are likely to be excluded from positions that have a direct, important role in determining the outcome of the contest. This has been applied to baseball pitchers, quarterbacks in football as well as place-kickers (Edwards, 1973; Johnson & Johnson, 1995; Eitzen & Sage, 2003). Edwards writes, “…the factor of centrality itself is significant only in so far as greater outcome control and leadership responsibilities are typically vested in centrally located positions since actors holding these positions have a better perspective on the total field activity” (1973, p. 209).

1.9       Coakley (2001) has asked why whites would be seen as more adept for leadership and control roles in terms of the United States race logic. This logic, or way of thinking about race in this country (US), manifests itself in many ways, particularly in how we ‘make sense’ of our social world. Essentially, there is an underlying stereotype in the US that African Americans are not ‘cut out’ for cerebral tasks, and this extends to the field of athletics. This belief, while not as overtly stated as in the past, still apparently lingers in the minds of many. The ‘dumb-jock’ stereotype – great athletically, but lacking intellectually – is still present, especially when used to describe African American athletes (Hoberman, 1997; Eitzen, 1999). For many years, the whites who controlled sport (coaches, general managers, etc) operated under the assumption that only whites were cut out for the so-called ‘thinking positions’ in major team sports (Coakley, 1998). As a result, African Americans have been pushed into certain positions and away from other positions. Positions of leadership and dependability have been filled with white athletes, while African American athletes are more likely to be found in positions of speed and power, with less of an explicit emphasis on cognitive qualities.

II. Purpose of the research

2.1       Although stacking patterns are not as rigidly enforced as they were years ago (in 1971 less than 1% of NFL quarterbacks were African American, and there were no African American centers), remnants of these patterns still remain. According to the NFL official web page, as of this writing (roughly one-third of the way through the 2002 National Football League season) there were 7 (out of 30) African American starting quarterbacks. Only two teams had an African American starting center. While 7 African American quarterbacks is surely an improvement over past years, it should be kept in mind that this represents only approximately 23% of the starting quarterbacks in the league. In comparison, 67% of all players in the NFL are African American, and they make up over 90% of all starting wide receivers, defensive backs, and tailbacks. 

2.2       Coakley (1998), drawing on research from Lavoie and Leonard (1994), Melnick (1996), and Melnick and Thompson (1996), maintains that there is no single reason for this type of racial stacking. Rather, it is based on numerous factors, including:

“1) Historical traditions related to ethnic relations in society…; 2) the history of a minority group’s involvement in a particular sport; 3) the proportion of minority-majority group members in a sport…; 4) the ethnic backgrounds of team coaches, general managers, and player scouts; 5) the degree to which different positions in a sport involve different skills and responsibilities; 6) the ways that positions are defined…; and 7) the perspectives used by those who identify and assess player skills and recruit players for teams” (p. 266).

This last factor mentioned in the above quote is the focus of the current paper. Specifically, how professional football scouts assess the mental and physical abilities of players in certain positions will be examined. Of particular interest is whether or not assessors of college players, when considering potential for the professional ranks, use different language when describing white and African American athletes.

2.3       The purposes of this paper then are twofold. The first relates to an assessment of whether scouting reports of college quarterbacks, centers, and inside linebackers, rely on mental descriptors of white players and physical descriptors of African American players. Following the aforementioned race logic, the research question suggests that descriptions of the athletes in these three positions, commonly held to represent ‘thinking’ positions, would be different for whites and African Americans, based solely on the ascribed characteristic of race. Namely, are white players in these positions described more in terms of their mental attributes, and, conversely, are African American players in these positions described more in terms of the physical attributes they are thought to possess?  The second purpose of this paper is a general one: to offer a valuable contribution to the literature on racial stacking and its possible underlying mechanisms in particular, and to the overarching sub-field of race and sport overall.

III. Methods

3.1       The procedure to test this research question consisted of analyzing the National Football League draft guides from the Sporting News magazine. According to their web page, the Sporting News is one of the oldest and most read sport magazines, with the number of ‘total adults reached’ at 2, 408,257. Starting in 1999, the magazine began publishing a special supplement to the magazine that was entirely dedicated to the NFL professional draft. Team needs were discussed, as well as the top collegiate players available for the draft at each position. A content analysis was undertaken to assess the descriptions of these players, in certain positions.  A total of 152 players were included in the analysis. A detailed description of the specific measures follows.

3.2       Race.   The race of each player was determined using a number of techniques. First, the magazine itself contained a few pictures of particular players. Therefore, the race was assessed for these players based on visual examination of the pictures in the magazine. However, pictures were not available for all players in the draft guide. For these remaining players, race was determined by one of the following means: in some instances, the author knew the race simply due to his knowledge of collegiate and professional football. That knowledge came from watching games, seeing interviews with stated players, reading stories with photos of the players, etc. If memory failed, players were looked up on the NFL web page. Since many of the players listed in the draft guide are now playing professionally, it was possible to find their photo on the roster section of each team’s home page.

3.3       There were a few players whose race could not be determined due to the unavailability of photos, or lack of knowledge about these players. Many of the players in the draft guide were not playing professionally, and were therefore hard to locate. In these cases, direct appeal to their former university was initiated. Simply, a representative in the athletic office of the particular school would be contacted, and asked the race of the former player in question. Using these methods, the racial background of each player in the study was attained. It should be noted that there is always a possibility of investigator error in determining race on the basis of photographs. However, this method has been widely used in the stacking research (Margolis & Piliavin, 1999; Jiobu, 1988; Best, 1987; Johnson & Johnson, 1995; Ball, 1973; Loy & McElvogue, 1970) and all efforts were made to ensure accuracy. Additionally, players whose race was ambiguous were left out of the analysis in a further attempt at precision. Therefore, the race of each player was as accurately defined, as the methodology would allow.

3.4       Position. Three separate positions were analyzed in the study. The first position was the quarterback. Traditionally, quarterbacks in college and professional football have almost exclusively been white. This began to change in the 1970s and has fluctuated since that time. Varying time periods have had as many as 4 or 5 African American starting quarterbacks or as few as zero at the professional ranks. The Sporting News draft guide used in this research has a description of each position, in terms of what the pro scouts look for in that position. Even though it is widely recognized that quarterback is considered a ‘thinking’ position, the description of the scouts, in terms of what they ‘look’ for, is included here, and will be for all three positions in the analysis (quarterback, center, and inside linebacker). This is done for reasons of clarity. Since the scouts are writing the descriptions of the athletes, it is important to understand what they are looking for. For example, regarding the quarterback, the position “needs the arm strength to throw the deep ball with velocity, the ability to throw to the sideline, the intelligence to recognize and exploit complex coverage and blitz schemes… Strong leadership and character are at a premium” (Sporting News, 2002 p. 20).

3.5       The second position considered in the study was that of offensive center. With the exception of the quarterback, the center might be the most important person on the offensive team. In addition to actually putting the ball in play by presenting it to the quarterback, the center is usually responsible for pointing out defensive formations to his teammates. In other words, the center usually makes sure that the remaining offensive linemen recognize the position and coverage of the defense, so that they may successfully anticipate the movements of the defenders. According to the Sporting News, “…the center must have good feet and be proficient as an angle blocker…though athletic ability and good feet are very important, size and strength cannot be ignored. Centers also need intelligence and football instincts. In many cases, they make all of the line calls and have to react quickly, in terms of anticipation and recognition” (Sporting News, 2000 p. 59).

3.6       The final position included in this study was that of inside linebacker. While the defense does not have as clear a signal and play caller as the offense (i.e. quarterback), the inside linebacker is the closest approximation to that role. In most defenses, the sideline coach will relay the play (either through hand signals or through a player) to the inside linebacker, who will then tell his teammates the proper defensive alignment. More important, this same player is usually responsible for recognizing any changes to the formation of the offense. Similar to the center, once lined up the inside linebacker needs to coordinate the other defensive players to make sure that all are in position. If the offensive formation shifts before a play begins, the inside linebacker is usually the defensive captain in charge of recognizing these changes and instructing his teammates accordingly. Inside linebackers “…are often described as quarterbacks of the defense. Along with physical skills, they must have great leadership qualities, intelligence and instincts” (Sporting News, 2001 p. 63).

3.7       A few notes about the coding of positions are in order. First, unlike previous studies on racial stacking in football, only the quarterback, center, and inside linebacker are included. Loy and McElvogue (1970) included offensive guards in their analysis as well as the positions listed above. Since they were examining the centrality hypothesis, guards were included as spatially central positions. Since the Sporting News draft guide does not describe the position of offensive guard in the same cerebral terms as that of center, this position was left out of the analysis. Ball (1973) included offensive guards also, as well as all linebackers. Even though he made some adaptations, his study was ultimately addressing spatial centrality, and in so doing, it makes intuitive sense to include all linebacker positions. Again, common knowledge, as well as the Sporting News draft guide descriptions, point to the more integral role of the inside linebacker in terms of command and leadership. For these reasons, only the inside linebacker has been included in this study.

3.8       A final note on playing position relates to the movement from college to the professional ranks. As mentioned earlier, some of the players listed will have switched positions once they reached the NFL. This could be the case with African Americans who played quarterback in college being moved to a different position such as wide receiver or defensive back. It could also be the case with the center position. A number of collegiate centers are moved to a different position, usually offensive guard, once they become professionals. Additionally, inside linebackers in college may or may not stay in the inside position once they reach the NFL. They may be moved to outside linebacker, defensive end, or safety. However, what is important for this study is their projected position in the National Football League, which was how they were addressed in the draft guides. The magazine took into account the players’ college position, as well as the position they would most likely play in the NFL, and generated their descriptions. This study makes no claim that a player described in one position in the draft guide would stay in that position for the duration of his professional career (if one were to ensue).

3.9       Physical Descriptions. A five-point Likert scale was used to rank the physical descriptions of the players in the above-mentioned positions. The points on the scale were as follows: a 5 would represent a player who possessed ‘excellent’ physical and athletic skills for the position in question; a 4 represents ‘above-average’ physical/athletic skills; a score of 2 corresponds with ‘below-average’ physical/athletic skills; and a score of 1 would represent ‘poor’ physical skills or athletic ability. The 3^rd point on the Likert scale could be one of two possibilities. Either the player was seen as having ‘average’ physical and athletic skills for the position; or, no mention was made of physical attributes. If a player was above average or below average on any particular scale, it would most likely be mentioned in the description. Therefore, failing to mention particular attributes is coded here as ‘average’.

3.10     The author carried out the positioning of each player on the physical scale subjectively. Each description, in terms of strengths and weaknesses, was read and processed holistically, in an attempt to gauge the overall description of the player. Certain terms and descriptions stood out and occurred often enough to be relatively easily grouped. For example, an ‘excellent’ score of 5 on the physical scale would include descriptions similar to this one (for a quarterback): “…Has a rifle arm and solid mechanics; is a powerful runner and threat in the open field”2. An example of an ‘average’ physical score of 3, for a center, might read, “…Has adequate strength for his size and is fundamentally sound”. Finally, for an inside linebacker, a ‘poor’ score of 1 on the physical scale would resemble the following:  “Does not show good hip movement, an rarely gets the proper depth on drops…lacks top speed and has marginal athleticism”. Following from this analysis, each player in the sample was assigned a score, 1-5, on a physical description scale.

3.11     Mental Descriptions. A five-point Likert scale was used to place each player in a continuum of perceived mental ability, and this was done in much the same way as the 5-point physical description scale. A score of 5 represented a player with ‘excellent’ intelligence or mental ability; a 4 showed ‘above-average’ intelligence; 2 symbolized a player with ‘below-average’ mental ability; and again, a score of 1 represented a player with ‘poor’ mental skills for the position. As with the physical description scale, a score of 3 on the mental description scale represented a player who was either average in his abilities, or it represented a player for whom no mental description was available.

3.12     A score of ‘excellent’ (5) for mental attributes, for an inside linebacker would read something like the following: “Extremely quick in reading and reacting to plays…understands where he needs to be on every play”. For a quarterback, a mental score of ‘average’ (3) would be similar to “…For the most part he sees the field and makes sound reads, but he has lapses and sometimes makes big mistakes in his reads”. Finally, a ‘poor’ mental score of 1, again for a linebacker, would resemble “…he has struggled in the classroom and off the field, and his instincts and intelligence have been questioned”. Using these techniques, a mental ability score was calculated for each quarterback, center, and inside linebacker.

3.13     Sample Description. A total of 152 players were initially included in the sample. For 7 of these players, it was not possible to discern their race, or, their race was classified as non-African American or non-white, and they were left out of the analysis. Those of Samoan or Hispanic descent, for example, were not included in the survey. The remaining 145 players consisted of 89 white players and 56 African Americans. There were 56 quarterbacks in the analysis, 36 centers, and 53 inside linebackers. Additionally, the positions in terms of race were broken down as follows: 41 quarterbacks were white and 15 were African American; 30 centers were white and 6 were African American; and 18 inside linebackers were white with 35 African Americans in this position.

3.14     The statistic used to analyze the physical and mental descriptions of the players was ordinary least squares regression (OLS). OLS was used because it predicts the amount of change in the physical and mental descriptions produced by the independent variable (the race of the player). The method estimates parameters by minimizing the difference between the observed response and the value predicted by the model (Everitt 1998). In other words, ordinary least squares attempts to find a straight line between the dependent and independent variables on which the sum of the squares of the errors is minimized. The initial statistical run (model 1) included all three positions in the analysis. Subsequent models included the quarterback position only (model 2), then the position of center (model 3), and finally the inside linebacker position (model 4). Each of these 4 models was done for both dependent variables- the physical description scale and the mental description scale.

IV. Results

4.1       Physical Description Scale. Table 1 presents the analysis of the physical description scale. Model 1 indicates that African Americans are significantly more likely to have higher physical description scores, for all three positions (b = .395, p < .001). In other words, whites are significantly less likely to be described in strong physical terms than African Americans, for all three positions analyzed in this study. This pattern was relatively consistent and robust for the remaining models examining each position separately. Hence, model 2, for quarterbacks, shows that African Americans received significantly higher physical ability scores than did their white counterparts in this position (b = .672, p < .001). The same pattern held for centers (b = .489, p < .001), with African American centers described stronger in terms of their physical characteristics. However, this pattern of results did not hold for inside linebackers, as there were no significant differences in the physical description scores between white inside linebackers and African American inside linebackers.

Table 1. Effect of Race on Physical Description Scale Scores

Variable	Model 1	Model 2	Model 3	Model 4
All 3 Positions	.395**
Separate Positions:
Quarterback		.672**
Center			.489*
Inside Linebacker				.162
Constant	3.256	3.390	3.000	3.333
Adjusted R²	.150	.451	.215	.026

Significance level: **p<0.001    *p<.01

4.2       How is this finding for inside linebackers to be interpreted?  One possibility would be the relative time of group occupation in this position. Of the three positions analyzed, the inside linebacker position has been ‘integrated’ the longest. Only until relatively recently have we seen more than one African American quarterback starting at any time in the National Football League. The position of center has been even slower to integrate. This being the case, it is possible that the physical differences between white and African American inside linebackers are not seen as crucial in the modern day, in terms of their ‘attractiveness’ as possible draftees. Indeed, it is conceivable that in positions that are least integrated, the focus is more strongly fixed on physical and mental differences between players. The explained variance (R-square) of the quarterback position is .441; for the center it is .215, and for the inside linebacker it is .026.  Therefore, the race of the player appears much more significant for the quarterback position than it is for the inside linebacker position. This gives at least tentative support to the proposed interpretation of the findings for the inside linebacker position.

4.3       Mental Description Scale. Table 2 presents the analysis of the mental description scale. Model 1 indicates that African Americans are significantly less likely to have higher mental description scores, for all three positions (b = -.441, p < .001). In other words, white players in the positions central to this study were significantly more likely than African Americans to be described in mentally strong terms. These results held true, for the most part, when the particular positions were examined separately. For example, model 2, for quarterbacks, shows that African Americans received significantly lower mental ability scores than did white quarterbacks (b = -.464, p < .001). This same pattern did not hold for centers however, as there were no significant differences in the mental description scales between white centers and African American centers.  Significance returns in model 4 for inside linebackers, with African American inside linebackers scoring significantly lower on the mental description scale (b = -.522, p < .001) than did white inside linebackers.

Table 2. Effect of Race on Mental Description Scale Scores

Variable	Model 1	Model 2	Model 3	Model 4
All 3 Positions	-.441**
Separate Positions:
Quarterback		-.464**
Center			-.205
Inside Linebacker				-.522**
Constant	3.802	3.780	3.778	3.889
Adjusted R²	.188	.201	.011	.258

Significance level: **p<0.001           

4.4       The discrepancy in the center position is most easily explained by the small sample size utilized within the study, particular in relation to the African American centers. This makes the interpretation of the center scores (both physical and mental) nebulous. As for inside linebackers, it appears that scouts are not as concerned with the physical skills listed as the mental ones. One interpretation of this could be that scouts are viewing African American and white linebackers as relatively equivalent physically; either whites are as athletic as African Americans in this position (following the expected direction suggested by the overall findings) or African Americans are less athletic at this position than others, thereby bringing their physical description scores closer to white scores. Conversely, despite the merging of the perceived athletic ability, scouts still perceive significant differences in the intelligence of the African American and white inside linebackers, with whites apparently possessing more mental ability for the position.

4.5       In summary, African Americans are more likely to be described by NFL scouts in positive terms for their physical and athletic ability than are whites. At the same time, they are less likely to be described in terms of their perceived mental or intellectual abilities. These findings held true for each position with the following exceptions:  there were no statistically significant differences between the ranking of the two races for the physical abilities of the inside linebacker position, nor the mental abilities of the center position.

V.        Discussion

5.1       The previous section reported the results of the analysis of racial differences in the mental and physical descriptions made by professional football scouts regarding the positions of quarterback, center, and inside linebacker. This final section discusses those results briefly and addresses the implications of these results for the sociological study of sport, as well as point out areas for future research.

5.2       Consistent with previous research and theoretical investigation, the results from this study suggest that white players in central (Loy & McElvogue, 1970), or outcome control positions (Edwards, 1973) within American Football, are more likely to be described in positive mental terms than are African American players in these positions. Conversely, African American players in these positions are more likely to be described positively in terms of their physical or athletic ability than are whites. What can be taken from this research?  At a minimum, this study should move the discussion on racial stacking forward, into relatively uncharted territory. With the recent increase in African American quarterbacks, there is a temptation to feel that racial stacking no longer exists. However, the findings here contradict this, hence, it remains important for the discipline of sociology to continue to bring attention to this situation. The issue of stereotypes regarding players of different races as a motivation for stacking has been addressed in a number of studies (Williams & Youssef, 1975), including this one, and this line of research is important for understanding the sociology of race and sport. The social psychological mechanisms involved in this process must continue to be studied.

5.3       The results of this study also point to the understood, ‘common sense’ beliefs about African American and white athletes. Even though African Americans have made tremendous inroads through the professional ranks in football, there are still lingering effects from the antiquated beliefs about their presumed mental inadequacies relative to whites. Fifty three percent of white Americans still feel that African Americans are less intelligent than whites (Turkel, 1992). For what reason should we believe that the sport world, which is a microcosm of US society, is immune from these racist attitudes? 

5.4       To make sweeping generalizations based on a study of this size would be premature. There is also the perfectly logical possibility that the descriptions of the scouts were accurate; i.e., African Americans ranked lower on most mental scores because the particular athletes in this study actually did fit these descriptions. Perhaps the athletic ability of white quarterbacks, for example, was significantly less impressive than African American quarterbacks, ‘across the board’. While this is entirely possible, the degree of significance and consistency of the findings renders this unlikely. Given the common stereotypes of African American athletic dominance, this finding is very likely to have been influenced by the subjective attitudes of the scouts, not just their objective skill determination.

5.5       Additionally, future studies would benefit from the addition of variables that would aid in assessing the actual veracity of the player descriptions, as well as other possible variables (as controls) that are not easily accessible. These possible intervening variables could include more detail about the athlete regarding individual sporting history, size, speed, strength, etc. Also, how much information from the scout is determined strictly by his/her observation of the athlete, and how much comes from other sources, such as other scouts, head coaches, assistant coaches, should be assessed. It would certainly be beneficial to include more draft guides, possibly from the scouts themselves or from the professional teams after the draft has taken place. What the scouts write, and how much of it actually makes the particular magazines may or may not mirror each other. Having more comprehensive descriptions, from multiple scouts, would aid in appraising the degree of stereotypes in the overall business.

5.6       A word about the study’s use of racial group membership is also in order. A limitation of this study is the use of theoretically homogenous racial groups, when in point of fact they are certainly heterogeneous. Since it is not possible to completely ascertain the absolute racial ancestry for players in the sample, a clear definition of racial group membership is elusive.  The definitions used here are very basic, and, as mentioned, were only determined by visual examination of photographs, in most cases. Future studies on racial stacking should address remedies to this common limitation of ‘racial’ research.

5.7       Regarding the veracity of the scouts’ reports, it would be helpful to gather additional data on the players’ statistics during their college careers. If a player is described as making bad decisions or being athletic and quick to the ball, is there a way to document whether or not these descriptions are accurate?  Finally, a longitudinal analysis would go a long way to describing possible trends in how African American and white players have been described. One would guess that 20 years ago, with far fewer African Americans in ‘thinking’ positions, scouting reports on particular athletes would be even more racially influenced. Whether that is true or not, and whether the situation is qualitatively different in today’s NFL, is unknown at this time. In conclusion, the results of this study should be seen as potential answers to exploratory questions, not as definitive ‘proof’ of system-wide bias. That said, the results do point in a particular direction; a direction that has been consistently pointed to for almost 35 years of research on racial stacking.

References

Ball, D. W. (1973). Ascription and position: A comparative analysis of “stacking” in professional football. The Canadian Review of Sociology and Anthropology, 10, 97-113.

Best, C. (1987). Experience and career length in professional football: The effect of positional segregation. Sociology of Sport Journal, 4 (4), 410-420.

Coakley, J. J. (1998). Sport in society (6^th Edition). New York: McGraw-Hill.

Coakley, J. J. (2001). Sport in society (7^th Edition). New York: McGraw-Hill.

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

Eitzen, D. S. (1999). Fair and foul: Beyond the myths and paradoxes of sport. Boulder, CO: Rowman and Littlefield Publishers, INC.

Eitzen, D. S. & Sage, G. H. (2003). Sociology of North American sport (7^th Edition). New York: McGraw-Hill.

Eitzen, D. S. & Sanford, D. C. (1975). The segregation of blacks by playing position in football: Accident or design?  Social Science Quarterly 55, 948-959.

Everitt, B. S. (1998). The Cambridge dictionary of statistics. Cambridge, UK:

Cambridge University Press.

Grusky, O. (1963). The effects of formal structure on managerial recruitment: A study of baseball organization. Sociometry 26, 345-353.

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

Hunter, D. W. (1998). Race and athletic performance: A physiological Review. In G. A. Sailes (Ed.), African Americans in sport (pp. 85-102) New Brunswick: Transaction Publishers.

Jiobu, R. M. (1988). Racial inequality in a public arena: The case of professional baseball. Social Forces 67 (2), 524- 534.

Johnson, B. D. & Johnson, N. R. (1995). Stacking and “stoppers”: A test of the outcome control hypothesis. Sociology of Sport Journal 12, 105-112.

Lavoie, M. & Leonard, W. M. (1994). In search of an alternative explanation of stacking in baseball: The uncertainty hypothesis. Sociology of Sport Journal 11 (2), 140-154.

Loy, J. W. (1975). Race and sport: A reaction to the McPherson and Madison and Landers papers. In D. M. Landers (Ed.), Social problems in athletics: Essays in the sociology of sport (pp. 157-167) Urbana, IL: University of Illinois Press.

Loy, J. W. & McElvogue, J. F. (1970). Racial segregation in American sport. International Review of Sport Sociology 5, 5-24.

Margolis, B. & Piliavin, J. A. (1999). “Stacking” in Major League Baseball: A multivariate analysis. Sociology of Sport Journal 16 (1), 16-34.

McPherson, B. D. (1975). The black athlete: An overview and analysis. In D. M. Landers (Ed.), Social problems in athletics: Essays in the sociology of sport (pp. 122-149) Urbana, IL: University of Illinois Press.

Melnick, M. J. (1996). Maori women and positional segregation in New Zealand netball: Another test of the Anglocentric Hypothesis. Sociology of Sport Journal 13 (3), 259-273.

Melnick, M. J. & Thompson, R. (1996). Maori people and positional segregation in New Zealand rugby football: A test of the anglocentric hypothesis. International Review for the Sociology of Sport 31 (2), 139-154.

Olsen, J. (1968). The black athlete- A shameful story. Sports Illustrated, July.

The Sporting News. (1999; 2000; 2001; 2002). Pro Football Draft Guide. New York: Times Mirror Magazine.

Terkel, S. (1992). Race: How blacks and whites feel about the American obsession. New York: The New Press.

Williams, R. L. & Youseff, A. I. (1975). Division of labor in college football along racial lines. International Journal of Sport Psychology 6 (1), 3-13.

Note:  I am grateful to D. Stanley Eitzen, Jay Trace Gilliam, Robyn Jones, and Steven P. Swinford for their willingness to review this work, and for the insightful comments they offered.