3

I wonder if it is fair is to compare the gini coefficient of two discrete distributions with different number of elements. If not, how can I adjust the coefficients for a fair comparison.

In particular, I'm interested in the case when the sum of values in both distributions are equal. For example, minutes played by players of two teams in a basketball game if they have different number of players (10 players in one team, 12 players in the other team) but for both teams sums plays 48 minutes x 5.

Other example, different partitioning of some surface in different regions, how to compare inequality for different partitioning if they have different number of subareas.

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
hernan
  • 61
  • 3

1 Answers1

0

Maybe you could start by calculating a standard error for the Gini coefs. This paper may help, it uses bootstrapping for comparison on data from Columbia. Another paper of interest, using monte carlo is Statistical Inference and Patterns of Inequality in the Global North

A similar question was asked here: Comparing Gini coefficients: Variance estimation etc. needed? (and so far, no answers).

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467