Ahh, Time to post this again...
http://www.sloansportsconference.com/wp-content/uploads/2015/02/SSAC15-RP-Poster-Paper-Referee-Analytics.pdf
Highlights:
"Using game data for teams in the six BCS automatically qualifying conferences (ACC, Big East, Big Ten, Big XII, PAC, SEC) from 2005-2012, we find clear evidence of officiating bias, most notably in the ACC and former Big East and, to a lesser degree, in the Big Ten and Big XII. Interestingly, evidence of bias varies between conferences. For instance, we found evidence that ACC officiating favors underdog teams, home teams, high prestige football programs, and long-time founding members of the conference. We found no evidence of this type or level of officiating bias in other conferences. "
"College football is unique among major U.S. sports because conferences govern how games are officiated, calling into greater question the oversight of in-house referees. Indeed, several studies suggest that unethical behavior is more likely with workers if their unethical actions further the goals of their organization [13, 14]. Recently, Business Week’s Bryan Gruley observed about conference referees, “But however pure their motives, however hard they work to keep bias out of their flag-tossing, these people are human...they know what the stakes are and where their bread is buttered” [15]. "
"The ACC, and to a lesser extent the former Big East, actually show evidence of the opposite bias: more favored teams are penalized to a greater degree"
"As with the game-level effects, the ACC shows more evidence of bias in its team-level effects than the other conferences. The ACC results demonstrate bias by penalizing teams with high football prestige less than their low prestige counterparts – a curious finding given the bias against the stronger football programs of the day. One explanation for this would be ACC officials protecting traditionally strong football programs while penalizing the more recently successful more. Supporting this notion, there is evidence of ACC officiating favoritism towards teams that have been in the league longest (founded in 1953) and more frequently flagging teams that are newer to the conference: Georgia Tech (1978), Florida State (1991), University of Miami (2004), Virginia Tech (2004), and Boston College (2005). Current ACC members Pittsburgh, Syracuse, and Louisville were all added after the sample period ended in 2012."
"Our findings support the inference that referees from the ACC in particular officiate in-conference games differently than referees in other major conference alliances. Table 1 shows an analysis of the Noll-Scully measure of parity [42], with the ACC showing the greatest parity in five of the eight sample years. Particularly given the betting line bias of the ACC in- conference, there is reason to believe that the ACC handicaps its stronger teams despite the financial incentives to do the opposite. Why does the ACC engage in this behavior and have these biases among officials? One possible explanation is the reputation of the ACC as a basketball conference with its four founding member North Carolina institutions (Duke, UNC, Wake Forest, and NC State) yielding the most political influence; internal ACC power may be threatened by non-founding schools with strong football that drive much of its revenue."