Discrimination in sports

Introduction

Discrimination based on religion, race, or ethnicity has been one of the US's major problems

and in professional sports. A significant amount of literature concludes that salary differences

exist among professional players in the NBA [ CITATION Law05 \l 1033 ] . Studies from the late

80s report that there was a significant difference in the salaries of players even though they

had the same level of productivity [ CITATION Koc88 \l 1033 ]. Studies have also shown that

viewership rates increase when there is greater participation by white players [ CITATION

Mar07 \l 1033 ]. These conclusions, made on the basis of research, are supported by extensive

data. However, a number of researches done after the NBA's globalization in 2006 have

concluded that no significant difference in salaries exists [ CITATION His17 \l 1033 ]. An

extensive amount of literature review points towards discrimination.

Discrimination in sports is an important aspect to study. It provides us an opportunity to

identify problems that should be countered to make sports more inclusive. Therefore, this

analysis aims to uncover the underlying factors which create the gap between salaries.

Data collection and methodology

Data used for this analysis includes different types of information, including player’s salaries,

player’s performance stats (points per game, games played), and player’s details (age, weight,

height). This data was collected using multiple sources, which include basketball-

reference.com and espn.com

The Salary variable has been transformed using the statistical standardization formula. The

variable White is a dummy which takes 1 for a white player and 0 for a player of colour. We

use this dummy to mainly identify discrimination.

are the variables added to incorporate the effect of player performance. Player’s details such

as age, experience, heigh and weight have also been included to increase the accuracy of the

regression models. Player’s positions are also available in the dataset.

Descriptive stats for the data

Media 25th
Variable name Mean St. Dev n Skewness Perc. 75th Perc.
Games Played 66.53 17.14 73.00 -1.81 59.25 78.00
Points per Game 14.63 5.45 13.90 0.55 11.50 17.50
Rebounds per
Game 5.55 2.72 4.90 1.28 3.70 6.78
Assists per Game 3.10 2.08 2.40 1.21 1.60 4.38
Steals per Game 0.97 0.44 0.90 0.69 0.70 1.18
Turnover per
Game 1.83 0.83 1.70 1.19 1.30 2.20
Age 27.39 4.52 27.00 -0.77 25.00 30.00
Exp 7.04 4.28 6.00 1.24 4.00 9.00
Height 78.77 7.09 80.00 -8.30 77.00 82.00
Weight 220.95 29.97 220.00 -1.74 202.75 240.00
Most Valuable
Player 0.04 0.20 0.00 4.74 0.00 0.00
Point Guard 0.20 0.40 0.00 1.52 0.00 0.00
Shooting Guard 0.20 0.40 0.00 1.52 0.00 0.00
Small Forwards 0.20 0.40 0.00 1.52 0.00 0.00
Power Forward 0.20 0.40 0.00 1.52 0.00 0.00
Centre 0.20 0.40 0.00 1.52 0.00 0.00
Allstar 0.17 0.38 0.00 1.74 0.00 0.00
White 0.28 0.45 0.00 0.99 0.00 1.00
Standardized
Salary 0.00 1.00 -0.11 0.46 -0.81 0.69

As mentioned earlier, the salary variable was standardized to remove the difference of scale.

The median for salary is lower than the mean which shows that it is positively skewed, also

confirmed by the 0.46 skewness coefficient. The skewness column shows the coefficient of

skewness. Negative number signals negative skewness, whereas positive number shows

positively skewed distribution. For instance, Number of games played has a negatively

skewed coefficient of 1.81 which means that the data for games played is negatively skewed.
Regression models and analysis

In order to make an accurate conclusion, I have used multiple regression models with varying

sets of variables. The first regression studies the impact of player performance and race on

the player’s salary. The dataset used for this regression is available in the excel workbook as

a sheet named Data for 1st regression.

Salary = Intercept + β1 (Games played) + β2 (Points per game) + β3 (Rebounds per game) +

β4 (Assists per Game) + β5 (Steals per game) + β6 (Turnover per Game) + β7 (Age) + β8

(Exp) + β9 (Height) + β10 (Weights) + β11 (White)

Regression results for this model show an r-square value of 73.4% which means that this

model is a good fit, but we have some variables with insignificant p-value. Coefficients for

games played, turnover and weight have a negative relation with salary. All other variables,

including White, have positive relation with the salary which shows that white players get

more salary as compared to the players of colour. Complete results are in the results sheet in

excel workbook.

The second regression was run, in STATA, on the same data and same variables but the

Experience variable was controlled to identify if there was a significant difference in the β10.

White had a coefficient of 1.544 in the first regression but when controlled for Experience, it

lowered down 1.469 which shows that Experience coupled with White leads to more salary

for a player. This model exhibits accuracy with an r-square of 79.9%.

Third regression was run with different set of variables. Salary is again used as a dependant

variable and player details (Age, Experience, Height, Weight) are used as independent

variables.

Salary = Intercept + β1 (Age) + β2 (Exp) + β3 (Height) + β4 (Weight) + β5 (White)

acceptable range. The coefficient for White is 1.755 and it is higher than the first two models.

Conclusion

The results obtained from the regressions show that white NBA players do get paid more than

their counterparts of colour. This conclusion conforms with the research studies of the past

that conclude that players are discriminated based on their ethnicity and other demographics.

The results highlight existence of discrimination in professional NBA and it links with the

economic theory of labour market discrimination. This economic theory is defined as

“unequal treatment of equally qualified workers”.

Kahn, L., & Shah, M. (2005). Race, Compensation and Contract Length in the NBA: 2001-

2002. Industrial Relations A Journal of Economy and Society.

Kanazawa, M. T., & Funk, J. P. (2007). Racial Discrimination in Professional Basketball:

Evidence from Nielsen Ratings. Economic Inquiry, 559-608.

Koch, J. V., & Hill, C. W. (1988). Is There Discrimination in the ‘Black Man’s Game’. 83-

94.

Naito, H., & Takagi, Y. (2017). Is racial salary discrimination disappearing in the NBA?

evidence from data during 1985–2015. International Review of Applied Economics.