Thank you for this reply! I still need to read the Subramanian piece but I will do that. However your explanation makes sense.

Your point is also taken regarding position rather than deprivation. However I still wonder about some of the implications of the measure. It is common for there to be large absolute differences in income between households/individuals at the very top of the distribution, as these people earn high incomes and an 1% difference in income, for example, is then a very large absolute deviation. Does an absolute measure not put undue influence on these differences, coming from a “left perspective”? Perhaps I can give an example of a hypothetical income distribution to make this clear.

Say there is a household in the bottom percentile of the income distribution, with an income of $1000 p/a. Then say there is a household in the 99th percentile, with an income of $1,000,000 p/a. Then say there is a household in the 100th percentile with an income of $3,000,000 p/a. I don’t think this is a wildly unrealistic income distribution.

Wouldn’t the absolute inequality measure suggest that we should care more about inequality between the households at the 99th and 100th percentiles, rather than that between the 1st and 99th percentiles? We go further in addressing income inequality by reducing reducing differences between households at the top than we do between households at the bottom and near the top? Is this not somewhat perverse from a left perspective?

LikeLike

LikeLike

I think Subramanian’s pieces linked above will be helpful to you on this question. But quick answer might be; if all parts of distribution were to grow in same proportional amount, a relative measure would show this change as constant inequality, even if absolute distance between the lower and upper ends were getting better. There could also be situations, like the US-India example above, in which a lower end of the distribution was growing faster proportionally, and so a relative measure of inequality would show declining inequality, even as an absolute measure showed increasing. So, basically, in most cases, an absolute measure would require a greater portion of income increases going towards the lower ends of distribution than a relative measure would in order to show a decline in inequality (in all cases maybe? I’d have to think about that).

As to the second part of your comment- remember we are talking about absolute differences here- not absolute deprivation- so it is still a measure that is about an individuals position related to others in society.

LikeLike

In general I am more used to left-wing people being attached to a view which takes relative position in society (and so-called “relative deprivation”) more seriously than conservatives, who tend to care only about “absolute deprivation”, if they care at all.

In any case, thank you for the thought-provoking post, I need to think about its implications a bit more.

LikeLike

Thank you for your comment. I definitely agree that there are many other important factors that go into determining the inclusiveness or not of growth, beyond any individual measure of overall inequality.

LikeLike

LikeLike