Research Paper Topics

Paul M. Sommers Spring, 2017

Suggested Research Paper Topics

1. Do teams (in the NHL, NFL, or NBA) raise ticket prices after successful seasons? Does fan interest rise after teams win division, conference, or league championships? Define a binary variable for each type of win: division, conference, and league. Does the fan’s taste for attending games rise, shifting the demand curve to the right?

[See R. M. Keohane and P. M. Sommers, “Win a World Series, Raise Ticket Prices.

But, Excessively?” Atlantic Economic Journal, vol. 36 (September 2008), pp. 381-

382.]
2. Who benefits from the presence of professional sports teams? Having a

professional franchise gives the city a higher profile which in turn attracts business

that would otherwise not locate there. But, real differences might not emerge unless

the city has a league-winning franchise. That is, there may be no noticeable improvement after the construction of a stadium or the addition of a team. How then does the business climate change in a city the year after one (or more) of its professional franchises wins a championship? [See Sports Illustrated, v. 108, no. 2,

January 21, 2008, p. 23 for a metric on a city’s championship(s).]
3. Historically, employers who integrated their baseball teams more quickly have

generally won more games than those who were slower to integrate. Gwartney and

Haworth (1974) underscored this result for Major League Baseball teams between

1952 and 1956. Can the same be said of baseball talent in Latin American or East

Asian countries? Or hockey teams that signed top players from former Eastern

Bloc countries?

pages 366-368.

standings were closer within a league over a season than when they were not? Use

the Gini coefficient or some other measure of competitive balance to measure the dispersion of team playing strengths.
6. Test the hypothesis that there is a positive relationship between NHL team violence (proxied by penalty minutes per game) and game attendance and, second, test the sub-hypothesis that violence is positively associated with American, not Canadian, spectators.
7. Is there ingroup favoritism in the NFL? Apostolatos and Sommers (2007) found

that there was evidence of ingroup favoritism in the NBA in the 2005-2006 season.

That is, in the NBA, teams with relatively more black players tend to win more

games if their head coach is also black. Two teams with black coaches represented

their respective conferences in the 2007 Super Bowl. So, one might (now) be

inclined to believe that there is a similar effect of the interaction between the racial

composition of a team and the race of the team’s coach on team performance in the

NFL. In summary, are predominantly non-white football teams more likely to win if their head coach is also nonwhite than if their head coach is white? More generally, are minorities in particular more likely to engage in behavior that furthers

organizational objectives if members of their minority group are represented in the

organization’s leadership? [See, for example, A.J. Apostolatos and P.M. Sommers,

“Is There Ingroup Favoritism in the NBA?” International Advances in Economic

Research, vol. 13 (February 2007), pp. 118-119.]
8. Do unequal salary distributions create morale problems severe enough to adversely

affect worker productivity? To determine whether or not the structure of

compensation in the NFL (or in the NBA) has any effect on worker shirking, regress winning percentage against the team’s Gini coefficient (for the team’s salary distribution) and average team salary. [See, for example, B.M. Avrutin and

P.M. Sommers, “Work Incentives and Salary Distributions in Major League Baseball,” Atlantic Economic Journal, vol. 35 (December 2007), pp. 509-510 and P.M. Sommers, “Work Incentives and Salary Distributions in the National Hockey League,” Atlantic Economic Journal, vol. 26, March 1998, p. 119. ]

distribution for a particular set of Topps baseball cards? That is, excluding all

common price cards, does there appear to be a premium for the cards of white

superstars (after controlling for player and team characteristics)? [See Hamilton

(1997) for a discussion of quantile regressions.]
10. If labor market opportunities in professional sports influence athletes’ graduation

rates, we would expect that the effect should be more dramatic in women’s

basketball after the emergence of the WNBA than it was before when there were

fewer professional opportunities in women’s basketball. [See DeBrock et al.,

1996 and E.R. Knopman and P.M. Sommers, “The WNBA and Graduation Rates of Women Basketball Players,” discussion paper.]
11. What has happened to the average length of job tenure in the NBA or NFL? Is it shorter now than it was ten or twenty years ago? In your paper, come up with a quantifiable measure of “team allegiance.” Compare “team allegiance” for athletes elected to that sport’s Hall of Fame in the 1990s to those elected in earlier decades.

[See P.M. Sommers, “Measuring Player Loyalty Among Baseball’s Hall of Famers,” Atlantic Economic Journal, vol. 35 (March 2007), pp. 125-126.]

12. For all pitchers (with a minimum number of IPs), regress their lifetime earned run average (ERA) against their years of experience in the majors. For a quadratic equation, there will be some point beyond which the ERA increases with additional years of experience. Has this point increased over time? That is, when you run this regression for data in (i) 1970, (ii) 1980, (iii) 1990, and (iv) 2000, you just might find that the point where d(ERA)/d(years) = 0 is rising over time. Still, in other words, the point where the ERA begins to rise with age/years of experience may

now be (much) higher than it was in the past. Why? Better conditioning? The use

of the DH? Steroids? [See P.M. Sommers, “The Changing Hitting Performance Profile in Major League Baseball, 1966-2006,” Journal of Sports Economics, vol. 9,

no. 4 (August 2008), pp. 435-440. In other words, I examined hitters in selected years across five decades. What are the results for pitchers?]

each card’s current price (see, for example, Beckett’s Baseball Card Monthly price guide, which tracks card values). Record each card’s price ten years ago. Regress the average annual rate of change in the card’s price against: (i) the player’s career batting and slugging average, (ii) a 0-1 variable denoting whether the ballplayer is nonwhite, (iii) a 0-1 variable denoting whether the card was a rookie card, and

(iv) a 0-1 variable denoting whether the ballplayer has been inducted into the Hall of Fame. [See C. Nardinelli and C. Simon, “Customer Racial Discrimination in the Market for Memorabilia: The Case of Baseball,” The Quarterly Journal of Economics, vol. 105, no. 3 (August 1990), pp. 575-595.]
14. Do new ballparks mean higher ticket prices? For each new stadium you decide to study, regress the club’s average ticket price (beginning five or ten years before the stadium opened) against a time trend variable; the club’s winning percentage the season before; a 0-1 variable that is zero before the stadium opened and “1” thereafter; and a 0-1 variable that is “1” if the new stadium was prompted by the arrival of a team from another city (and zero, otherwise). [See The Sporting News, April 4, 2000 at http://tsn.sportingnews.com/baseball/articles/2000404/228259.html and http://rodneyfort.com/PHSportsEcon/SportsEcon.htm for data on average ticket prices.]

15. Does a professional franchise give a metropolitan area a higher profile, which in turn attracts business that would otherwise not locate there? For metropolitan areas with more than 750,000 persons (see Table 661 in the Statistical Abstract of the United States 2008), regress the annual percent change in personal income against the population of the metropolitan area, the unemployment rate, and three dummy (0-1) variables to denote the presence of a football team, baseball team, and basketball team in the metropolitan area.

and after their trades.

U-shape fit reach a minimum? Is this value for an age between 25 and 29?

than teams that do not make the playoffs? If possible, examine every year since

free agency, since one could argue that it was impossible for players to exercise

loyalty prior to free agency. [See P.M. Sommers, “Measuring Player Loyalty Among Baseball’s Hall of Famers,” Atlantic Economic Journal, vol. 35 (March 2007), pp. 125-126.]

relative to the remaining teams in their league? Have MLB teams with new stadiums won more games per season than teams without new stadiums? Take 1-, 3-, and 5-year averages before and after. What is the experience in other sports leagues? Include and then exclude expansion teams. Is the attendance and winning percent increment a function of the public’s support funding the new stadium? [See Rodney Fort, Sports Economics, 2^nd edition, Tables 10.1 and 11.1, pages 340 and 388.]

35. How badly has Tiger’s departure from the PGA Tour affected viewership of PGA

golf tournaments he has missed in 2009-2010? Assuming that you can get information on the number of viewers from The Nielsen Company, compare (in a paired t-test) the number of viewers at PGA tournaments he missed relative to the year before. What has been the average percentage change in viewership at these tournaments? Has viewership been less adversely affected by his departure now than it was when he missed part of the 2008-09 Tour due to injury?

36. The WNBA was founded in 1996 and their inaugural season was 1997.  The NCAA has published online the graduation success rates (GSRs) by sport and by gender for all D-I schools for the last five years.  The first year of data (2005) actually includes students who entered college in the years 1995 through 1998 (most of which are pre-WNBA).  The last year of data (2009) includes students who entered in the years 1999 through 2002 (all of which follow the WNBA's inaugural season in 1997).  For the term paper, I would like you to: (i) run a series of paired t-tests on GSR rates between 2005 and 2009 for all D-I schools by major conference and

(ii) run a series of paired t-tests for the women's "Sweet 16" teams in 2006 (compared to their average GSRs in 2005), the "Sweet 16" teams in 2007 (compared to their average GSRs in 2006), and so forth.  

My hunch is that you will find (with few exceptions) that GSRs are, on average, no lower (or higher) after the inception of the WNBA than before.  But, there may be a few conferences (in relatively poor parts of the country) where the women could not resist the temptation to leave college early.

Famers before and after free agency in any of the three other major professional team sports.

lacrosse (men’s or women’s) or a national championship in any varsity sport

(i) increases the percentage of alumni that give to Middlebury College,

(ii) increases the size of the gift per donation, or (iii) increases the average

SAT scores of entering freshmen? Divide alumni into men and women. If you

use dollar amounts, remember to express changes in real terms.

39. What matters more to franchise values – population or income? Or, are there

regional differences in franchise values that have nothing to do with either the

size of the market (as measured by the population) or income? For this topic,

use the most recent Forbes estimates of franchise values in the NFL, MLB,

the NBA, and the NHL. For each league, are franchise values higher the greater

the population/income in the area? For which sport is the relationship most (least) pronounced? For each league, compute the compound annual rate of appreciation for each team using data for 2005 as the base year. Determine whether those rates of appreciation are correlated with the annual population growth rates in the franchise’s hometown over the same seven-year (2005 to 2012) period.

significantly higher in games at the beginning of a home stand? Are they

significantly higher if the series goes to a game 7? Are they significantly higher

if there’s at least one large market team in the series? Do ratings depend on the

population of the NL/AL team’s hometown? On the combined population of the two teams?
41. In the aftermath of Lance Armstrong’s admission to doping, first develop an economic model of deterrence that focuses on the economic incentives and

economic consequences of cheating. (See, for example, Roger Blair’s discussion

of “Cheating in Sports,” chapter 11 in his Sports Economics, Cambridge University Press, 2012.) Apply the model to the sport of cycling, indicating what the benefits

are as well as the punishment if the cyclist is caught. Provide several numerical estimates of expected wealth under various assumptions about the probability of detection and different penalties for cheating.

“early entrant” is an NCAA player with remaining eligibility or any player who has been out of high school for at least three years.) Pro-Football-Reference.com uses a comprehensive metric, created by mathematician Doug Drinen, called Approximate Value (AV) to assess player worth at any position in any year. Compare the average AV in their rookie year of early entrants to all other players drafted early, say, in (i) the first two rounds and (ii) all subsequent rounds. Repeat for the 2008, 2010, and 2012 NFL drafts.

interested in minimizing travel cost, both time spent on the road and gasoline

cost. For the given constraint that one game must be played on the road and one game must be played at home against each opponent, find the “ideal” schedule for all ten teams which minimizes their collective (that is, NESCAC’s) travel cost.

How does the “ideal” schedule depart from each school’s 2012-13 actual schedule

against other conference teams?
44. In Major League Baseball, the Baseball Almanac reports that as of May 2012, there

have been 30 suspensions (issued to fewer than 30 ballplayers due to repeat

offenders) since the performance-enhancing drug policy was put into place in 2005

(see “Steroid Suspensions,” Baseball Almanac at http://ww.baseball-almanac.com/legendary/steroids_baseball.shtml ). For these ballplayers, compare

their average stats before and after the suspension. In most cases, did the player

actually perform significantly worse after the suspension?

45. Has the distribution of prize money in tennis become more highly skewed for men than for women? Find the Gini coefficient for the total winnings of the top

100 men and women players for each of the last twenty years. Regress the Gini

coefficient against a time trend variable, a separate regression for men and for

women. Is the slope coefficient in each case discernible from zero? That is,

has inequality on the women’s (or men’s) tour changed over time?
46. Some researchers have concluded that conference standings and national rankings in college football have become increasingly predictable. (See, for example,

E. Woodrow Eckard, “The NCAA Cartel and Competitive Balance in College

Football,” Southern Economic Journal, vol. 72, no. 4, pp. 826-845.) For each

of the six largest conferences – Southeastern Conference, Big Ten, Atlantic Coast Conference, Pac-12, Big 12, and Big East – regress σ_w against a time trend

to assess this claim over (at least) the last ten years.
47. Are unrestricted free agents (UFAs) in the NHL more likely to perform worse after

signing a multi-year contract with their same team than are UFAs who sign (a multi-year contract) with another team? Or, is their average number of points per game about the same in both their contract year and the year thereafter? Examine

UFAs in each of the last five years.
48. Will the Sochi Olympics be the last Games featuring NHL players? NHL owners claim that when the NHL league schedule is suspended for the Olympics attendance is strongest and that, according to Kevin Allen of USA Today, “some markets have had trouble re-acquiring their fan base after the shutdown.” For each NHL team, was their a discernible difference in average home attendance two (or three) weeks before and after the Sochi Games? Was this also the case following the Vancouver Games? How might you use regression analysis to test this claim?
49. Use http://geomidpoint.com/calculation.html to find the geographic midpoint of the men’s

and women’s “Sweet 16” basketball teams for each of the last twenty years. Use the same

methodology in Sommers’ “Mathematics and Geography: The Changing Geographic Midpoint of Baseball’s All-Stars” to gauge movement in the midpoint

over time.

career length of other NFL players at their position? Examine players (and player reps) from the year 2000.

Baseball” (Institute for the Study of Labor, IZA, Discussion Paper No. 5695, May 2011), authors Francesca Cornaglia and Naomi Feldman argue that married

ballplayers receive, on average, higher wages than single ballplayers. Is there a

marriage premium for athletes in any of the other three major sports leagues?

Is there a fatherhood premium?
52. Which teams have been most successful (over, say, the last ten years) in the NFL

Draft? For each season, one could look up each team’s draft history, that is, the

players drafted and their career approximate value (CarAV). See, for example

the Minnesota Vikings, www.pro-football-reference.com/teams/min/darft.htm.

Is the average CarAV for one team equal to the average CarAV for another?

A series of two-sample t-tests would indicate if one team has been more successful

than another (in each of the last ten seasons). A table listing the accumulated CarAV for each team each season alone would be of interest, as would be a table

that lists the average CarAV for all picks by round over the last ten years.

NFL players drafted between 2002 and 2014. For all 224 picks, regress CarAV

against the pick number. Now, use the estimated CarAV of all 224 picks to

evaluate actual trades of draft picks made by NFL teams over the last five or ten years. Do you find that teams that typically play in the postseason have made

“better” trades than teams that typically do not play in the postseason?
54. There were 182 reported concussions from 2015 regular-season games, a 58 percent

over the corresponding number in 2014. Is there any pattern to the number? Were

there more reported in the second half of the season than in the first half? In the

second half of the game than in the first half? More from AFC games than NFC

games? Disproportionately more these last two seasons from some teams than from

others? More from rushing plays than from passing plays? More in games played

in southern states than from northern states? Was the game (in which at least one

concussion occurred) decided by eight points or less?