Documenting the #RedskinsPride Disaster

Yesterday, the Washington Redskins organization asked their fans to tweet @SenatorReid using the hashtag “#RedskinsPride” to tell him what the team meant to them.  There are now numerous examples of how firms should let hashtag campaigns develop organically rather than try to encourage a conversation over which they have no control.  However, the Redskins organization decided to ignore common sense, and the results were predictable.

The chart below tracks the hourly Twitter mentions of “#RedskinsPride” on its primary axis, and the hourly sentiment of all tweets containing the hashtag on its secondary axis (The sentiment is indexed from 1-100, with 100 being the most positive).

#RedskinsPride

While the chart provides a nice overall view of what happened yesterday, it is interesting to see how the hashtag campaign evolved over time and geography.  The table below describes the evolution:

#RedskinsPride Chart

We should note that while a lot of the tweets came from outside of the Metropolitan DC area, even the tweets originating from DC, VA, and MD tended to be more negative than positive.

Manish Tripathi & Mike Lewis, Emory University 2014.

More on NFL Fan Equity: Dynamics & Mascots

Last week we presented our ranking of NFL fan bases.  The Cowboys, Patriots, Jets and Saints headed this list, and every other city in America let us know that our study was garbage.  As in any study of this nature, there will always be limitations that leave room for debate.

One such source of debate is in how much data we use for assessing fan equity.  We use 11 years of data to develop a model for forecasting expected consumer demand (the forecast is based on winning percentage, pricing, stadium capacity, metro area population, metro area median income and other factors).  We then determined fan equity (fan loyalty and support) by comparing the model forecasts to each team’s last three years of results.

One important question is whether three years is sufficient.  In our minds three years is a compromise.  An argument in favor of a lengthier time horizon is that fan loyalty is a persistent trait that moves slowly.  If this is the case, it might make sense to look at relative performance for the last five or ten years of data.  On the other hand, the world is constantly changing and evolving so it also makes sense to focus on recent history.  In the case of sports, if championships and post season success are the sources of long-term fan equity then using a shorter horizon that is sensitive to near term changes makes sense.

The top five for the last the last decade would be New England, Washington, Kansas City, Denver and Pittsburgh.  This list will likely make other fans happy but it will still result in significant unhappiness in Green Bay.  Later this week we will discuss in more depth why some of the teams that conventional wisdom would suggest to be at the top of our list fell short.

We can also look at who is rising and who is falling.  For this analysis we compare the fan equity rankings using the first 3 years of the data (2002 to 2005) with the last three years (2010 to 2012).  The analysis finds that the biggest risers were the Cowboys, Jets and Colts.  The four biggest drops were the Chiefs, Buccaneers, Rams and Redskins.  This list shows both the pros and cons of using short versus long horizons.  The short horizon allows us to capture the long-term impact of what Peyton Manning delivered the Colts and the importance of the Cowboys’ new stadium.  On the negative side, the early success of the Rams and the Bucs seems to have turned out to be short lived.

There is one other element of the preceding list of teams that have suffered a decline in fan equity that may raise some eyebrows.  Two of the teams that suffered dramatic drops have Native American oriented team names: The Chiefs and the Redskins.  Over the last decade we have witnessed an increase in efforts to eliminate Native American team names and mascots.  Lewis is an Illinois grad and Tripathi is a Redskins fan so they know firsthand how a mascot controversy can split a fan base.  There are, of course, alternative explanations for why these two teams’ fan equity has decreased (but keep in mind that we do control for team performance) but it is at least noteworthy that two of the four teams with the biggest drops have controversial team names.

Mike Lewis & Manish Tripathi, Emory University 2013.

An “Imperfect” Analysis of the Economics of Native American Mascots: Much Ado about Nothing?

A few days ago I came across an article in Forbes that discussed the continuing controversy surrounding the Washington, D.C. NFL franchise’s use of “Redskins” as a team name.  As a University of Illinois graduate and fan, the issue of Native American mascots is something that hits close to home.  As many may know, the University of Illinois dropped the “Chief” in 2007 while keeping the Illini name.  There are many Illinois fans that view the decision with derision, and think of it as political correctness gone wild.

One nice consequence of the Emory Sports Marketing Analytics (ESMA) project is that we have assembled a significant amount of data that provides opportunities to address many topical questions.  The issue of how Native American Mascots affect a team’s revenues and brand equity is (almost) one such question.  I say almost because while the team name currently in the news belongs to an NFL franchise, the most relevant data we have available comes from NCAA basketball.

Here’s the upshot of our study examining the impact of changing the Native American mascot: Schools experience a very short (1 or 2 years) negative financial impact and then quickly recover.  Furthermore, in the long-term, the shift away from a Native American mascot yields positive financial returns. Now, an examination of our study, and all of the caveats associated with it.

NCAA basketball includes numerous examples of teams that have dropped, kept or adapted Native American mascots.  For example, St. John’s and Marquette have dropped Native American mascots and changed the team nickname.  Illinois and Bradley have retained their team names but dropped all Native American imagery.  Other schools such as Florida State and Utah have received permission from tribal representatives and made no changes to their mascots or team names.  But, it is not perfect data because there are only a relatively small number of examples where teams change names (and as noted above there are a large number of special circumstances such as the Seminole tribe’s embracing of FSU).  The approach we have taken to analyzing the impact of Native American mascots and switches away from using a Native American mascot therefore requires some creativity and several assumptions.

Our first analysis involved predicting men’s basketball revenues as a function of a wide variety of factors including: season winning percentage, tournament success, stadium capacity, enrollment, historical successes (championships, final fours, tourney appearances), and conference affiliation.  For this analysis, we also included variables related to the school’s history and current status with regard to Native American mascots.  Specifically, we included a variable that indicated if the school had ever had a Native American mascot, the time since a change had been made and whether the school had kept a Native American mascot (for this analysis we treated Bradley and Illinois as having dropped the Native American mascot).  This analysis yielded a marginally significant negative term for the variable that indicated if the school had ever had a Native American mascot, and a significant positive term for the time since a shift away from a Native American mascot.  (We should also note that as always we experimented with a large number of alternative specifications, such as log transforms of time since shift and quadratic formulations.  The findings we report are robust in that they occur across these various specifications).

In terms of financial impact, the model results suggest that school’s experience a very short (1 or 2 years) negative impact and then quickly recover.  The results also suggest that in the long-term the shift away from a Native American mascot yields positive financial returns.  As a follow up, we used the brand equity measures created here as a dependent variable and regressed this value against the previously defined variables related to the school’s use of a Native American mascot.  In this analysis we found NO significant effects.  The key implication is that switching away from a Native American mascot has no long-term negative effect on brand equity.

The preceding analysis requires a fair amount of caveats.  For instance, it may not be reasonable to assume that findings from college basketball translate to pro football.  College fans often refer to their teams based on the name of the institution (Illinois beats Maryland) while pro fans might more often refer to team names (The Bears crush the Redskins).  It should also be noted that the data available for studying the effects of mascots is limited.  For example, there are only three schools that have received permission to use tribal names.  This is too small of a sample to make conclusions about the effect of tribal permission on the outcomes generated by Native American mascots.

All that said, we believe that our findings have a great deal of face validity.  While the use of Native American mascots is controversial, it is not clear to us that the consequence of a shift away from a Native American mascot is that big of a deal.  While some fans may complain, it is not clear that these fans actually change their behavior or their shopping habits.  It might also be that merchandise sales become more appealing to segments that did not like the previous Native American mascot.

Mike Lewis & Manish Tripathi, Emory University 2013.