How is Bruce Bartlett Cheating: Let us Count the Ways

Robert Waldmann

Steve Benen made a graph out of a table from a column by Bruce Bartlett. Bruce Bartlett is an interesting figure — a heterodox conservative who praises Reagan and criticizes Bush Jr. The figure sure fits Bartlett’s line. It shows the effective tax rate on families with median income. Sad to say, Steve Benen’s web page is a *.mht ??? and I don’t know how to steal the figure [see UPDATE below] so just click.

Bartlett identified the “effective federal income tax rate — taxes paid as a share of income — for a family with the median income. The median is the exact middle of the income distribution — half of families are above and half are below. It’s as close as we can get, statistically, to the typical American family.”

[UPDATE: I do know how to steal graphics, but you should read the Benen piece as well. Graphic above by Steve Benen, from data presented by Bruce Bartlett. -klh]

It shows an increase up until the election of saint Ronnie, then a decrease, a sharp decrease in 1986, more decline when the GOP took control of Congress, a down tick when Bush was elected and Bartlett still supported him, then a gradual increase as Bartlett became a critic.

Oh how convenient. I describe how he is cheating after the jump.

As anyone who has ever filled out a 1040EZ must know, this is nonsense. Effective tax rates do not depend only on income. The table would be meaningful if Bartlett had calculated the average effective tax rate of families with the median income, but that’s not what he said. I suspect that he used a “tax family” with median income and characteristics such that the graph looked the way he wanted.

Of course, Bartlett is talking about the income tax only. The increase in FICA under Reagan doesn’t appear. That is par for the course for a Republican.

Less importantly for the median family, he doesn’t consider the capital gains tax. However, the average family with median income pays positive capital gains tax (the average of a non negative variable must be positive).

I’m guessing that Bartlett’s “median family” has 2 earners and 2 children. I see a big drop when the marriage penalty was reduced.

The average family with median income has less than 2 earners and less than 2 children. The family which pays the median effective tax rate … I have no idea.

I will now check my guess (starting 5:59 EDT) by clicking a link.

6:00 AM EDT. Yep he lied. The table which appears when I click his link is Entitled “Average and Marginal Income Tax Rates for Four Person Families …”

That is not “as close as we can get statistically to the typical family.”
The Bureau of the Census projected in 1996 that that in 2005 the average number of persons per family would be 3.09 which is rather less than 4.

It does seem that I was wrong about the number of earners. They assume one earner not two. Median income is for four person families.

Now part of what is going on here is that the “Tax Policy Center” is trying to make Reagan look good. However, that isn’t all that’s going on. They care about what they consider to be normal families. Single adults without children and single mothers with children don’t count. Women are assumed to be housewives. The only people that matter are those in families which consist of one working dad, mom and two children. For families with median income single mothers with three children slip into the sampel (lucky duckies). For half median income they don’t as it is assumed that there are 2 children in the family for the purposes of the EITC.

I think their prejudices about what is normal is even stronger than their Hackitude as Bush Jr would look better if one considered the marriage penalty.