More on the Dividend Tax Cut

The picture first appeared in Henry Waxman’s study of Bush’s proposed dividend tax cut, then appeared on gorilla-a-gogo, which Off the Cuff then linked and copied, which CalPundit then also posted ,with attribution. If you still haven’t seen it, here it is (for those unclear on the regressive concept, I’ve identified which bar applies to you, unless you make over $374k per year):

This graph is a great example of why I was earlier able to state with confidence that “[Grassley’s] proposal is not Bush’s plan; it’s just the best he [Bush] can get without Snowe, Voinovich, Collins, and Chafee (Zell Miller can only sell out once). Were Bush’s plan implemented, I assure you it would be regressive.”

How would things look under Grassley’s plan? The rich would get to exclude $500 of dividend income, which would otherwise be taxed at 38.6% (2003 rate), for a savings of $193; the bottom 80% would get an average saving identical to that in Waxman’s graph, $29.50. Even that figure overstates the benefit–the $29.50 stems from a relatively small number of upper middle class (70-80%) getting a benefit well above $29.50 and the majority of the middle and lower class getting 0. Seriously, would it be so hard to identify the mean and the median benefit of the various tax programs? In this case, the median benefit to those in the bottom 80% would be the benefit to the person in the 40th percentile, which I suspect would be very close to zero–maybe $2.50 in tax savings.

Whenever the average of X is well above the median of X, you know that the distribution of X is skewed upward. Suppose you and ten friends are in a room and that the average income in the room is $50k, which is also the median (half make more than $50k, half make less). Now take your wealthiest friend and replace him with Bill Gates, who makes $1 billion per year. The average income becomes $104.5 million, but the median remains unchanged at $50 thousand. More to come.


P.S. Why all the emphasis on attribution? See the previous post.