As we have insisted on our website, white fans support non-white athletes more then they do white athletes. Here is an excerpt from a scholorly article by C. Depken and J. Ford, titled:
Customer based discrimination against MLB Players: additonal evidence from All-star ballots
website location is: http://www.uta.edu/depken/P/cbd.pdf#search='ethnic%20origin% 20of%20baseball%20players'
"The empirical results show that, even after controlling for player and team attributes, fans indicated a bias in favor of Blacks and Latinos during the sample period.13 As seen in the Model I, Latinos received approximately 65,000 more All-star votes than Whites of equal characteristics, ceteris paribus. Confirming Hanssen and Andersen (1999), we find no evidence of customer-based
discrimination against Blacks. Rather Blacks were favored over Whites as they received approximately 52,000 more votes than Whites, ceteris paribus. Model II, in which the Blacks and Latinos dummy variables are combined into a single dummy variable called non-white (NWHITE), supports the finding of customer-based discrimination in favor of Latinos and Blacks. The estimated coefficient on NWHITE suggests that non-white players received an average of 58,000 more All-star
votes than Whites during the sample period, all else equal."
Customer based discrimination against MLB Players: additonal evidence from All-star ballots
website location is: http://www.uta.edu/depken/P/cbd.pdf#search='ethnic%20origin% 20of%20baseball%20players'
"The empirical results show that, even after controlling for player and team attributes, fans indicated a bias in favor of Blacks and Latinos during the sample period.13 As seen in the Model I, Latinos received approximately 65,000 more All-star votes than Whites of equal characteristics, ceteris paribus. Confirming Hanssen and Andersen (1999), we find no evidence of customer-based
discrimination against Blacks. Rather Blacks were favored over Whites as they received approximately 52,000 more votes than Whites, ceteris paribus. Model II, in which the Blacks and Latinos dummy variables are combined into a single dummy variable called non-white (NWHITE), supports the finding of customer-based discrimination in favor of Latinos and Blacks. The estimated coefficient on NWHITE suggests that non-white players received an average of 58,000 more All-star
votes than Whites during the sample period, all else equal."