Boston University Professor Todd IdsonEC385 Economics of Sports Spring 2018, Midterm #1SolutionsInstructions: Answer all questions below in your blue books (be sure to show all of your calculations). Please label all parts of your diagrams and draw them large enough so that all aspects can be readily assessed. 1. Omitted variable misspecification is when a variable is not taken into account, i.e. included in the regression, but it does affect the dependent variable. This will bias the estimated effect of any included variables that are correlated with the omitted variable. The example focused on in this paper is estimating an attendance function. Turnover was always ignored in the earlier models when we estimate attendance function. However fans care about roster stability, i.e. form attachments to players.Attendance is negatively correlated with turnover. Thus, turnover should belong in the regression. So firstly, the author measure the degree of roster turnover (different ways to measure this, but they opt for weighting by salary to give name players a greater weight/importance) and tests whether they have a statistically and quantitatively significant effect on attendance. As a result, the coefficient of turnover is negative and significant. Then author pointed out, if we omitted the turnover variable, it will cause misspecification problem here: (1) winning percentage has a positive effect on attendance, (2) roster turnover has a negative effect on attendance, and (3) winning percentage and roster turnover are negatively correlated. As a result, if you omit roster turnover from the attendance regression, the effect of roster turnover on attendance will be partly reflected in the estimated winning percentage effect of turnover, acting to upwardly bias the estimated effect of winning percentage on attendance. To see thisnote that “low” turnover yields higher attendance, and low turnover is associated with higher winning percentage, and higher winning percentage leads to higher attendance, i.e. part of the estimated positive effect of winning percentage on attendance will also be reflecting the fact that when winning percentage is “high” turnover tends to be low which in itself is exerting a positive effect on attendance. If turnover was included in the regression, then the estimated winning percentage effect would fall. More specifically, consider simple estimation model as follow:True model: Misspecification: Note that is negative, is also negative, and is always positive. i.e. is positive and it means that the coefficient of winning percentage is upwardly biased if turnover is omitted in the regression.2. Answers: Not necessarily. The restriction will break a vertical integrated firm into separate upstream and downstream components. This may actually worsen the well-being of the consumer because the downstream firm will buy at an inflated monopoly price (P up below) from the upstream firms and treat this as the marginal cost and then price accordingly, charging P down. If it was vertically integrated, i.e. one firm in the bottom diagrams, then the media outlet would cost the games at marginal cost (i.e., P up would be treated as the MC at the second stage) and hence charge a lower pricefor the broadcast.3. Answer: If the Bills practice second degree price discrimination, they can effectively charge a price equal to the value of the eight tickets, i.e. the sum of the WTP for each ticket, P = 135+120+105+90+75+60+45+30 = $660 for an 8-game packageThis would allow the team to capture all of the consumer surplus.4. Answer: The owner can claim a depreciation expense of $40 million per year ($200 million/5). The reduction of profits by $40 million, which would have been taxed at a rate of 40 percent, saves the owner $16 million per year, or $80 million over the five year
View Full Document