BULLETIN! The course of American democracy may be decided on the first Saturday of November in the following college football games: Florida vs Georgia, Michigan vs Indiana, Illinois vs Ohio State, and Texas vs Texas Tech.
That is, if you buy the fascinating notion, revealed at a recent academic conference, that college football outcomes -- win or lose -- perhaps can sway gubernatorial elections by as much as 10 percent.
At the Citadel Symposium on Southern Politics in March, some of my political science colleagues presented research demonstrating, statistically, the possible influence of football victories on the fortunes of incumbent governors and parties.
Considering that some of the most important gubernatorial campaigns of 2014 are occurring in states like Florida, Michigan, Illinois, Ohio, and Texas -- with serious consequences for national politics and the 2016 presidential election -- perhaps we had better look closely at their research and its ramifications for American democracy.
This intriguing proposition comes from Keith Lee (a doctoral student at the University of Florida), Sydny Bryan (an MPA student at Valdosta State University), and Dr. James LaPlant (a professor at Valdosta State University). They examined the relationship between major college football teams' victories/losses before general elections and votes for the incumbent party in the local county of those teams -- for all state contests in the period 2010-2012. Their datataset covered gubernatorial elections across 41 states (if a state did not have an NCAA Division 1A football team, it was not included in their dataset).
And their findings were astounding.
Football Victory = 5 percent to 10 percent Electoral Boost?
According to Lee/Bryan/LaPlant:
Our findings from the most recent gubernatorial election cycle reveal a more noticeable effect of college football victories the weekend right before the election. In our regression models of the incumbent party vote share in gubernatorial elections, the total number of football wins in the two weeks before the election is the most powerful predictor.
They report a direct, statistical connection between winning the football game on Saturday and local electoral support on Tuesday: "In counties where the college football team won the weekend before the election, the incumbent party vote share is 10 percentage points higher than where the team lost (54.11 percent vs 43.59 percent)." When they controlled for key demographic and economic factors, "a victory in the two weeks before the election produces approximately a 5 percentage point increase in the incumbent party vote share."
Of course, there was plenty of light-hearted banter among the presenters and attendees at this scholarly gathering. And the authors were not predicting winners and losers in statewide elections based on specific games and home-county returns.
But their explanation and data are really quite reasonable. In many areas, football is king; and game day is a frenzied celebration of tradition and pride. In some places, the game is a cultural religion, a mystical ritual of who we are and what we hold sacred. It is not a far stretch to understand that "the game" means as much to some folks' feelings of well-being as does the state of the economy; and losing a "big game" right before the election seems to fit in with such ideas as the impact of rainy weather on one party or another. Certainly, in close elections, any depressing news triggers concern among ruling regimes and might exert a small effect on turnout and margin of victory or loss.
Interestingly, the presenters told us that the influence of grid-iron victory does not extend to all aspects of political life. Their research showed that professional football records did not impact vote share in mayoral elections; and they found no relationship between college football records and the 2012 presidential election. Apparently, the authors note, "as elections move closer to the people, the impact of college football outcomes becomes more evident."
Is This a Southern Thing?
Lee/Bryan/LaPlant originally explored the football-elections relationship because of their interest in the possible dynamics of regional culture. The South epitomizes an environment where football sometimes assumes fanatical stature; and, as most national observers know, the Southeastern Conference boasts magnificence on the field and in the minds of the southern faithful. However, the authors found that the overall relationship was mixed when they compared the South to other sections of the country:
For the gubernatorial elections from 2010-2012, the impact of college football wins 10 days before the election on incumbent party vote share is more pronounced in southern counties when compared to non-southern counties. For the weekend right before the election, the impact of college football victories is slightly more pronounced in non-southern counties. In our regression model, the interaction term of the South x total football wins in the two weeks before the election is positive but statistically insignificant. Clear regional differences are simply not evident in this study.
My colleagues candidly acknowledged that their presentation was unusual and might reflect poorly on American democracy. However, they boldly asserted the essential validity of their research: "In the last cycle of gubernatorial elections, our findings provide further evidence that voter well-being and happiness can influence retrospective voting." They will have to do further research to figure out whether they're on to something about college football and electoral victory; fortunately, they said that they are pursuing that course.
I have spent most of my life as a political scientist and public official; so I approach such propositions with due professional caution. But, early in my career, I was a sports writer for the Atlanta Journal; and I have personal experience with football fanatics. Thus, I'm not shocked at the seeming connection between winning games and winning elections.
I cannot wait to see whether the first Saturday in November impacts the 2014 elections.
AUTHOR NOTE: This column is part of a series of posts about Southern Politics. These posts derive from the 2014 Citadel Symposium on Southern Politics, a gathering of regional specialists in historic Charleston, SC. This Symposium has been held every-other-year since 1978; and it has become a "main event" for serious South-watchers from around the country. A hundred specialists--representing scholars from about fifty academic institutions--participated in the most recent conference, March 6-7, 2014. In this series, I will attempt to incorporate pertinent aspects of the presented papers and some of my own comments into various themes.