NBA Salary Inequality for the 2016-17 Season

Now that the NBA playoffs are well underway, I thought I would come back to looking at salary inequality among NBA players.  As I wrote earlier, NBA salary inequality had been improving over the last few years.  So I again I collected the NBA salary data from ESPN and calculated the Gini coefficient for the league as a whole.  For the most recent season, we see that salary inequality has taken a step backwards, with the Gini coefficient now 0.5519 and is worse than the US Gini coefficient for 2014 of 0.480.

NBA Salary Income Inequality since 2005

Now that the NBA season is well underway, I thought I would look at how equal (or unequal) NBA salaries are over the last few years.  To do that I collected the NBA salary data from ESPN and calculated the Gini coefficient for the league as a whole.  Results are listed in the table below by season, the number of players (n), the Gini coefficient and the total salary for all players in the league.

Season	n		Gini		Total
2005-2006	394		0.4917		$1,619,154,926
2006-2007	409		0.5006		$1,697,883,876
2007-2008	415		0.5161		$1,817,073,513
2008-2009	413		0.5053		$2,002,564,767
2009-2010	428		0.5260		$1,972,670,023
2010-2011	450		0.5366		$1,964,987,905
2011-2012	501		0.5500		$2,013,125,239
2012-2013	541		0.5601		$2,062,073,682
2013-2014	391		0.5409		$1,786,979,523
2014-2015	431		0.5294		$1,903,052,600
2015-2016	407		0.5017		$2,126,853,179

As you can see the NBA has been fairly unequal in the distribution of salary among its members, but over the last few years, salary inequality has been decreasing in part reflecting the increase in average salary over the last few years.  For comparison purposes, the US Gini coefficient for 2014 is 0.480.  Below is a chart of the Gini coefficient in the NBA since the 2005-2006 season.

2015 NBA Payroll and Performance

Today I am going to look at NBA payroll and NBA performance.  Over the last few years, NBA payroll has not been a very good predictor of NBA regular season performance.  Let's see if that trend holds up for the last two NBA regular seasons.  Thus I have grabbed data from the internet on NBA payrolls  for 2013/14 regular season and the 2014/15 regular season and NBA performance and run a regression on how well relative payroll is related to regular season performance.

What I find is that over the last two NBA regular seasons is that relative payroll is positive and statistically significant, but only explains about 12.6% of regular season winning percentage.

2014-2015 NBA Competitive Balance

At the end of the regular season last year I blogged about competitive balance in the NBA since the 99/00 season.  Now I will update this given the end of this year's regular season.  For those interested here is a guide to calculating competitive balance using the Noll-Scully method on your own.  The only difference here is that I will give both the sample and population measures of the Noll-Scully for this season.

After downloading the data from basketball reference, I found that the NBA is still highly uncompetitive as compared to other sports leagues, with a Noll-Scully of 2.97 (sample) and 2.92 (population).  Either way, the NBA is still very uncompetitive relative to other sports leagues.

2015 NBA Attendance Analysis

Now that the NBA regular season is in the books, let's take a look at how home attendance has changed (or not changed) over the last two seasons using the NBA attendance numbers from ESPN.  So specifically I want to know if average regular season home attendance has changed from the 2013/14 season to the 2014/15 season statistically.  I choose to use a t-test (assuming unequal variance) in NBA teams average home regular season attendance.  What I find is exactly what I found previously, and that is that NBA average home regular season attendance has not changed in a statistical sense.  In this case, the value of the t-test for NBA average home regular season attendance between the two seasons is 0.215, which is greater than generally accepted confidence level of 0.05.  Hence, I conclude that NBA average home regular season attendance in the 2014/15 season is statistically no different than in the 2013/14 season.

NBA Attendance Since the 2011/12 Lockout

Has average home attendance changed since the lockout?  In our book, The Wages of Wins, we analyze season attendance in the NFL, NBA, NHL and MLB with and without player strikes/ owner lockouts and conclude using intervention analysis that strikes or lockouts do not have a significant difference in attendance for each of those four US sports leagues.  About two years ago I blogged about NBA attendance right after the lockout and also concluded that average home attendance was not different during the lockout year as opposed to average home attendance before the lockout.

Now that we have two more years of attendance data, I thought that I would return to this topic of NBA average home attendance again.  To do so I am analyzing the two seasons before the lockout, the lockout season and the two seasons after the lockout.  As I did in the blog linked in the paragraph above, I will use a t test (assuming unequal variance) on the data that I collected from ESPN's NBA attendance webpage.  As a reminder, I will be analyzing average NBA home attendance since the lockout and non-lockout seasons have different total games played and thus different total attendance.  I am not interested in whether total attendance is different, but whether fans are going at different rates when the games are played.  For those interested in total team attendance, I have included this at the end of this blog.

Running the numbers results in the same conclusion:  average NBA home attendance is not statistically different in either post lockout season compared to average NBA home attendance of the lockout year.  Statistically average NBA attendance is not different when comparing the lockout year to the two seasons before or after the lockout.

Here is a chart of the average home attendance data since the 2009/2010 NBA season.

In terms of total home attendance, there is a statistically significant difference in the year before the lockout and the lockout season and also comparing the lockout season and the season after using a t test assuming unequal variance.

For those interested, here is a chart of total NBA home attendance for those five seasons.