Saturday, May 13, 2006

Reich on Taxes

I often enjoy reading Robert Reich, former Clinton Labor Secretary and now Berkeley professor. In the new 4th edition of my principles text, I reprint one of his articles. (If you're curious, it's an article on manufacturing employment, reprinted in the chapter on Earnings and Discrimination). But over at his new blog, Reich reports some "facts" about the current tax law that I found so surprising that I decided to check them out.

Here is what Reich says:
Some administration apologists, including the editorial page of the Wall Street Journal, claim repeatedly that the rich are paying a larger-than-ever share of income taxes, so it’s entirely fitting that they get the lion’s share of any tax cut.

This logic conveniently leaves out two facts. First, the rich are now paying a smaller percentage of their income in taxes than at any time in the last seventy-five years. That they pay a lot of taxes nonetheless is a by-product of the mind-boggling increase in their income and wealth relative to most other Americans. Second, if you consider not just income and capital-gains taxes but all the taxes people pay – including payroll taxes and sales taxes – you find that middle-income workers are now paying a larger share of their incomes than people at or near the top. We have turned the principle of a graduated, progressive tax on its head.
The best place to look to check these alleged facts is the Congressional Budget Office website. The CBO has one of the best staffs around, and it has a long history of being nonpartisan. The CBO regularly calculates average tax rates by income group.

Here is the Total Effective Federal Tax Rate for 2005, according to a CBO report (Table 2):

Lowest quintile 5.5 percent
Second quintile 12.0
Middle quintile 15.6
Fourth quintile 19.6
Highest quintile 26.3

Top 10 percent 27.8
Top 5 percent 29.0
Top 1 percent 31.1

These data (which include all federal taxes, not just income taxes) seem hard to square with Reich's second claim that "middle-income workers are now paying a larger share of their incomes than people at or near the top."

Reich's first alleged fact is that "the rich are now paying a smaller percentage of their income in taxes than at any time in the last seventy-five years." To evaluate this claim, I compared that 31.1 percent number for the richest 1 percent with historical tax rates, available from CBO here (Table 1A). CBO begins their series in 1979.

What I found is that the average tax rate for the top 1 percent was below the current 31.1 percent rate for 10 of the 11 years from 1982 to 1992. It reached a low of 25.5 percent in 1986. For better or worse, President Bush has not come close to reducing tax rates for the rich to the levels they achieved after President Reagan cut taxes.

One might be curious how the tax rate on the lowest quintile--5.5 percent in 2005--compares with historical data. The CBO data show that this tax rate has fluctuated since 1979 between 5.2 to 10.2 percent. Federal taxes on the poor at now near the low end of their historical range.

I don't know what data Reich was using when he made his claims. He does not post a link to a data source at his blog. If there are data backing up Reich's claims, it would be interesting to figure out why they are inconsistent with CBO data. Comments and explanations from readers are welcome.

---

Updates for data geeks

#1: Brad DeLong suggests that the CBO numbers for 2005 are imperfect because they fail to include AMT relief that was passed subsequent to the CBO analysis. A fair point. He then suggests looking at 2004 numbers instead. Here I am inclined to disagree, because the 2004 figures include the effects of temporary bonus depreciation. Neither year's data are ideal if we want to evaluate where we are right now.

In any event, Brad's point is moot. For the sake of argument, let's use the CBO figures for 2004 instead of 2005. One could then defend Brad's weaker claim that "President Bush has reduced tax rates for the rich to roughly the levels they achieved in the Reagan years." Nonetheless, it remains the case that neither of Reich's more striking "facts" appears defensible on the basis of CBO data.

#2: Phill Swagel helpfully points me to supplemental CBO tables where the effects of temporary bonus depreciation are removed from the data. Using this supplement, and Brad's preferred year 2004, we find (in Table B2) these average tax rates:

Lowest quintile 5.3 percent
Second quintile 11.3
Middle quintile 14.8
Fourth quintile 18.8
Highest quintile 25.2

Top 10 percent 26.7
Top 5 percent 27.9
Top 1 percent 30.0

The story using the corrected 2004 data is largely the same as I described in my original post using 2005 data.