Question

The following appeared in the Wall Street Journal, September 16, 2019 "Letters to the Editor" that refers to a September 11 article that appeared in the Journal. Read the letter reproduced below in part and answer the question following the letter. "Phil Gramm and Mike Solon start “Warren’s Assault on Retiree Wealth” (op-ed, Sept. 11) by telling the reader that the households of ages 65 to 74 have an average of $1,066,000 in net worth. This may be technically true but it has little significance in a political or public-policy context. The (much more meaningful) median figure for the same age group is $224,000—less than one quarter of the figure they cite....." Tim McGlinn Maplewood, N.J. Which of the two measures of net worth, $1,066,000 from the original article, or the letter writers measure, $224,000, do you think is the superior measure to use? Explain fully and specifically.

Answer 1

An arithmetic average of any data set assumes that each data point has equal weightage. In the context of net worth of a household, it implies that all the households have an equal amount of money. This is obviously not true. Wealth distribution is highly skewed in reality. Bill Gates, whose age is 63, would alone contribute Billions to the total sum while a lower middle class household's net worth would hardly affect it. Thus the statement that the average net worth is $1,066,000 just shows that the total sum of all net worths is quite huge. But it doesn't reflect on the distribution of the wealth.

Median of a data set tells us the value which lies around midway in terms of magnitude. In our context, a median net worth of $224,000 means that around 50% of the households' are below this value and the remaining 50% lie above it. Thus the median value isn't swayed by the extremely high or extremely low values.

Going by the above two paragraphs, it can be convincingly argued that the median net worth is a superior measure to use. It shows that the middle most household and all those above it have at least $224,000. No such conclusion can be drawn from the MEAN value.