Lewie Pollis is a senior at Brown University and a former baseball analytics intern for the Cleveland Indians. He also writes for ESPN Insider. Follow him on Twitter @LewsOnFirst.

Two things can be said about Barry Bonds’ place in baseball history: He was one of the greatest players the game has ever seen, and he may have been the single most controversial player of all time. He earned the former title largely by setting MLB records with 73 home runs in 2001 and 762 in his career (as well as 166.8 WARP), while the latter distinction can be traced to his having done all that under the cloud of allegations that he used some cocktail of performance-enhancing drugs.

But a close look at public opinion polling suggests another partial explanation for why opinions of Bonds are so strong and divided: race. Not only do divisions of opinions on Bonds exist along racial lines—they do so just as some political science literature would predict.

The question of how race has impacted popular perceptions of Bonds is not a new one. By 2006, as Bonds was closing in on Babe Ruth on the all-time home run list, there was significant debate about whether Bonds’ generally negative public image stemmed from the color of his skin—as he claimed—and how much of it came from his PED use or treatment of the media. Even the guys at Fire Joe Morgan weighed in on the issue. Ken Tremendous claimed that Bonds “claimed racism everywhere he went for whatever reason if it suited his purposes,” while dak asked the provocative question: “Do you think he would have booed so vehemently if he had been white?”

Before we go further, allow me to say that discussions of race in academia are different than they are in the mainstream media or casual conversation. Political scientists in particular are very matter-of-fact about questions of race and the existence of racism. It is also worth emphasizing that the trends they describe for different demographics are generalizations and do not apply to every individual person within those groups.

There are three specific findings from political science research that I believe are important for contextualizing the numbers that will follow. (The works listed here are not necessarily the only writings on the subject. I use these as examples because they illustrate the points well. I would highly recommend reading them in full, as the full publications are far richer and more interesting than I can convey in a couple sentences.)

  • In Obama’s Race: The 2008 Election and the Dream of a Post-Racial America, Michael Tesler and David O. Sears argue that Barack Obama’s visibility during the 2008 campaign and beyond polarized voters along racial lines regarding not just Obama but any issue with which he was associated. If their model is correct, Bonds’ prominence throughout the 2000s would have primed fans to think about him (and perhaps the broader questions of PEDs and morality in baseball) through the lens of race. (Full disclosure: Tesler is my concentration advisor at Brown.)
  • We can infer from the survey data presented by Paul M. Sniderman et al. in Reasoning and Choice: Explorations in Political Psychology that manifestations of modern racism are heterogeneous across people’s conformances to social norms—i.e., that the difference between how a racist would see two upstanding citizens of different races is not the same as the difference between how he or she would see two criminals of different races. This means that the effect of race on public perceptions of Bonds would likely be different than it would for another, less-controversial black player.
  • In Race and the Formation of Attitudes: Responses to Hurricane Katrina, Lonna Rae Atkeson and Chiere D. Maestas find evidence that members of a demographic group are more supportive of one of their own if they feel the issue in question is divisive along demographic lines. This means black fans could be more supportive of Bonds if they inferred racial undertones in the criticism of him than they would be if race were a non-issue.

With that in mind, how could race have affected public opinion on Barry Bonds?

The trend can be seen first at the most basic level: personal favorability. In a Gallup/USA Today poll from February 2009, only 17 percent of self-identified white respondents expressed a favorable opinion on Bonds compared to 55 percent who viewed him unfavorably; other respondents (full racial breakdowns were not easily available for most of the polls I looked at) had a slightly positive view of him, 30-28. This difference is statistically significant, while no significant racial differences existed for Roger Clemens, Alex Rodriguez, and Derek Jeter in the same poll. Other surveys reveal that this is a consistent trend. In a December 2004 Gallup/CNN/USA Today poll on whether celebrities (including Bonds) had been “naughty” or “nice” that year, whites broke for naughty by a nearly two-to-one margin (52-27) while other respondents were evenly divided (40-39)—another statistically significant difference.

When it came to allegations of drug use, whites were less likely to give Bonds the benefit of the doubt across the board. In an ABC News/ESPN poll in July 2006, a whopping 80 percent of white respondents thought he had probably used steroids against just 55 percent of other respondents. A 2004 Gallup/CNN/USA Today poll revealed that only 15 percent of white baseball fans believed Bonds on the issue of PED use, against 26 percent of other fans; five years later, in a 2009 Gallup/USA Today survey, only eight percent of white respondents said they believed that Bonds had taken steroids unknowingly while 22 percent of nonwhites were willing to give him the benefit of the doubt. And white fans were also far more likely than other fans to think that Bonds was being treated fairly regarding the steroid allegations against him, both in a 2006 Associated Press/America Online poll (whites: 71-18, other respondents: 54-35) and in a 2007 Associated Press/Ipsos-Public Affairs poll (whites: 63-24, other respondents: 49-38). (All of the racial differences cited in this paragraph are statistically significant.)

How would we explain this in terms of the political science research? There are plenty of reasons besides race to dislike Bonds on a personal level or to think he used PEDs. But the statistically significant differences between how whites and other respondents see Bonds make sense in the context of Tesler and Sears’ research: as a highly visible black celebrity, he may have activated racial resentment among fans. If that is the case—a theory supported by the racial differences in perceptions of whether or not the media had been treating Bonds fairly—Atkeson and Maestas’ findings suggest that black respondents would support him out of solidarity, which could partly explain why he is seen in a positive light among nonwhites despite his reputation as a disagreeable person whom people of any race could dislike on a personal level.

Whites were also the harshest respondents in terms of how baseball should deal with Bonds. A March 2006 Gallup/CNN/USA Today poll revealed that most self-identified white fans did not think Bonds should be in the Hall of Fame (40-51) while a two-thirds majority of other fans endorsed him for Cooperstown (67-26); though the gap narrowed in both directions in a June 2007 Associated Press/Ipsos-Public Affairs poll (whites: 47-42, other fans: 59-31), a significant difference remained. More strikingly, a July 2006 ABC News/ESPN poll revealed that nearly half of whites (47-51) thought Major League Baseball should immediately suspend Bonds after he was charged with perjury—a position supported by less than a quarter of other respondents (23-73). (All of the racial differences cited in this paragraph are statistically significant.)

As Tesler and Sears observed with Obama, specifically naming Bonds adds a racial component to public opinion that is not otherwise there. In a February 2009 CBS News/New York Times poll of baseball fans, there was no significant difference between what whites and other respondents thought should be done with records set by steroid users. Yet when a July 2006 ABC News/ESPN poll specifically asked about Bonds’ records, less than a third (31 percent) of whites said they should remain on the books as normal, compared to more than half (56 percent) of other respondents. And while race did not significantly impact whether baseball fans thought players who used steroids before the game officially banned them in 2002 should be allowed into the Hall of Fame in an Associated Press/Ipsos-Public Affairs poll in April 2005, when Bonds was specifically named in an Associated Press/America Online poll a year later whites were evenly divided (48-48) while other fans supported him two-to-one (66-33).

The Hall of Fame question also allows us to directly test Bonds’ uniqueness in dividing fans along racial lines. A December 2006 Gallup/USA Today poll asked baseball fans about the Hall of Fame worthiness of nine candidates: Bonds, Mark McGwire, Cal Ripken Jr., Tony Gwynn, Jim Rice, Goose Gossage, Jose Canseco, Rafael Palmeiro, and Sammy Sosa. Only Bonds registered a statistically significant difference between how white fans and other respondents viewed him. That Gwynn’s support was virtually identical across racial lines makes sense in light of the inferences to be drawn from Sniderman et al.’s research that the effects of racial resentment would be different for a universally well-respected black player than one who is already mired in controversy.

But perhaps the most interesting result is the response to the question of whether or not Bonds should be considered the legitimate home run king, given that he used PEDs to do it, or if Hank Aaron still held the record in the minds of those polled. An NBC News/Wall Street Journal poll in June 2007 revealed that while a plurality of blacks thought Bonds deserved the honor (37-34), white respondents registered as almost four-to-one against (15-59). That such a dramatic racial difference exists is striking considering that Bonds and Aaron are both black. This fits with the implications of Sniderman et al.’s findings: that race matters differently in public opinion for someone who is already unlikable than it does for someone who is universally respected. It is also possible that this discrepancy can be explained by Tesler and Sears’ model of racial priming—Bonds has been in the public eye far more than Aaron in recent years, so his race is more easily accessible than Aaron’s—but given that the survey question specifically mentioned both of them it is not necessarily safe to assume that race is significantly more salient for Bonds than for Aaron.

Finally, we have a reasonable counterfactual for what public opinion on Bonds might look like if he were not black in Roger Clemens, another all-time great whose legacy has been tainted by PEDs (and who’s had an equally tough time gaining admittance to Cooperstown). According to the survey data, race has had no impact on perceptions of Clemens. There was no significant difference between what whites and other respondents thought of Clemens’ Hall of Fame candidacy or of him personally in a February 2009 Gallup/USA Today poll that found statistically significant racial gaps for Bonds, nor did race matter in a March 2008 CBS News poll on whether or not Clemens should be eligible for the Hall of Fame. What else could cause this phenomenon for Bonds but not Clemens if not Bonds’ race?

It should be noted that these numbers might be outdated. All the polls of interest that I could find were at least five years old, and things could have changed since then, especially as Bonds has been less in the public eye (leading to fewer opportunities to prime racial judgments) and the anger about his PED use may have started to subside (making him less unlikable for non-racial reasons). I would guess that race still influences opinions of Bonds today, though not to the extent that it did at the peak of the Bonds-related controversies.

It is impossible to precisely measure the impact of Bonds’ race on how the public perceives him, whether due to racial resentment from whites or supportive solidarity from blacks (and sympathetic fans of other races), and certainly there are reasons to dislike him on a personal level that have nothing to do with race. But as we begin to consider Bonds’ legacy, we should take caution to ensure that history’s judgments are not based on the color of his skin.

All survey data courtesy of the Roper Center Public Opinion Archives.