Quote of the day: Lambert and value judgments

By Matt Nolan 02/11/2013


I was excited to see James post about value judgments this morning – as that is exactly what I was about to throw a brief post on!  Partially motivated by this:

But also motivated by the fact I’ve been reading a bunch of ‘normative economics’ recently.  Here in the book “The Distribution and Redistribution of Income” by Peter Lambert is a quote about value judgments (with reference to, in this case, income inequality measures)”

It is hard to avoid making (often well-concealed) value judgments when assessing inequality

The points he goes on to make regarding valuing income distributions given certain measures are relatively well known, but worth repeating:

  1. The Lorenz curve only implies we value relative income rank (not absolute, or non-income factors)
  2. The Gini coefficient is a representation of this curve, and values transfers over equal distances of the rank of income – so it will value a transfer of $1 from someone rich to someone a little rich in the same way as it will value a transfer of a $1 from someone poor to someone a little more poor.
  3. The Schutz coefficient is another representation of this curve.  However, it only values transfers from above the mean to below the mean – and it values them the same way irrespective of rank.

These “objective measures” both hide considerable value judgments, and judgments that when put out in the open seem inappropriate.  Hence why economists go on to talk about more general principles of social welfare, and how to compare income distributions given that.

The goal of economic description should be largely to make assumptions transparent.  And as James hints at towards the end of his post, we should be careful about letting social convention within economics convince us that bad assumptions are actually good ;)