How Progressive Are Our Taxes?

Stephen Dubner has a post on the “Freakonomics” blog titled “The next time someone tells you that taxes are not progressive…” He relays information from a new Congressional Budget Office (CBO) report, via Greg Mankiw, which lists effective federal tax rates for households at various points in the income distribution. The rates are higher for those with larger incomes. The implication is that our tax system is quite progressive.

But it doesn’t make much sense to look only at federal taxes. State and local taxes account for about a third of total tax revenues, and they tend to be less progressive than federal taxes.

If we take into account all taxes — federal, state, and local — the effective tax rate for the well-to-do is only a bit higher than for the poor. Here is one way to see this, based on data from the CBO and the Tax Foundation.

21 thoughts on “How Progressive Are Our Taxes?

  1. It is also interesting that when certain Serious Economists or Right Thinking Experts argue about cutting taxes, they always argue first about cutting capital gains, then corporate and income taxes, and I have not seen one Serious Economist or Right Thinking Experts advocate cutting sales or payroll taxes.

  2. I think the percentage paid by the very wealthiest groups, top 5%, 1%, 0.1% would probably go down, with greater access to tax shelters and deductions, and with the capped and non-progressive pay-roll tax going to zero as a percentage of earnings.

    Plus, capital gains taxes are much less than income taxes. EPI’s The State of Working America 2006/2007 has on page 79 a table showing that the top 1%, in 2006, got 33.2% of it’s income from capital income, 20.4% from business income, 11.1% from “other income”, and only 35.3% from wage and salary.

    Look at what Justin Wolfers had to say in a May 1st post:

    Do the well-off pay their fair share, or do they also deserve a tax break?

    Well, let’s start with the ultra-rich. Bajillionaire Warren Buffett has argued that he isn’t being asked to pay his share. He went around his office, asking people what share of their income they pay in income taxes. Buffett’s 17.7 percent tax rate compared a bit too favorably with the 30 percent tax rate paid by his secretary.

    So it appears that the tax system favors the super-rich over working stiffs.

    And Buffett went a step further, putting his money where his mouth is. Last November he issued a challenge to his fellow billionaires:

    I’ll bet a million dollars against any member of the Forbes 400 who challenges me that the average (federal tax rate including income and payroll taxes) for the Forbes 400 will be less than the average of their receptionists.

    So far, no-one has taken him up on this bet.

  3. At the website there is an article “A Final Report Card on the Reagan Years?”. The short story is the top .01 percent of taxpayers have an income of $412 billion (11,000 households making over $8.6 million) versus the bottom 20 percent with an income of $385 billion (24 million households with income below $17,000). A person with $10 million dollars isn’t any happier than a person with $9 million dollars. Kit

  4. Lane,

    I would love to see a similar analysis to the one Mankiw provides that incorporates state and local tax rates, but it is very difficult to actually to see this in the graph.

    Do you have a more clear depiction of the total effective tax rates for each quintile? Breaking it down into tax payments and pretax income kind of loses the issue in the margins. Percentages would be more powerful.

    Also, CBO used total taxes by total income, would that be the same for this graph?

  5. Another way to look at it is Figure 3 from the Tax Foundation document you cite. If you include federal and state taxes, the system is still progressive, but notably less so.

    Federal, State and Local Effective Tax Rates, Calendar Year 2004
    By Income Quintiles

    1: 13%
    2: 23%
    3: 28%
    4: 31%
    5: 35%

  6. Steve, thank you. Those numbers appear to be the proper counterparts to the Mankiw numbers.

    For comparison:

    Federal / Federal/State/Local rates
    Lowest quintile: 4.3 percent / 13%
    Second quintile: 9.9 percent / 23%
    Middle quintile: 14.2 percent / 28%
    Fourth quintile: 17.4 percent / 31%
    Percentiles 81-90: 20.3 percent / (Fifth quintile mean: 35%)
    Percentiles 91-95: 22.4 percent/ (Fifth quintile mean: 35%)
    Percentiles 96-99: 25.7 percent/ (Fifth quintile mean: 35%)
    Percentiles 99.0-99.5: 29.7 percent/ (Fifth quintile mean: 35%)
    Percentiles 99.5-99.9: 31.2 percent/ (Fifth quintile mean: 35%)
    Percentiles 99.9-99.99: 32.1 percent/ (Fifth quintile mean: 35%)
    Top 0.01 Percentile: 31.5 percent/ (Fifth quintile mean: 35%)

    I am presuming that the numbers would start off lower than 35% for the smaller percentiles Mankiw slices up for the fifth quintile, and end near if not higher than 40% for the top .01 percentile.

    I suppose we can only speculate there.

    But great numbers nonetheless!

  7. Steve and Publius,

    Actually, the Tax Foundation calculations of effective rates that Steve cites aren’t directly comparable to the CBO ones, because the Tax Foundation adjusts the incomes of each quintile for the estimated cost of transfers (see Appendix B of the working paper at if you’re interested in the details). Using Tax Foundation numbers comparable to those of the CBO (but including state and local taxes), the effective tax rates for the five quintiles in 2004 are:

    Q1: 12.4% (bottom quintile)
    Q2: 21.0%
    Q3: 24.9%
    Q4: 27.2%
    Q5: 29.4% (top quintile)

    But note also that both the CBO and Tax Foundation include government transfers in the incomes used to calculate effective tax rates. There’s nothing illegitimate about this per se; but if we want to see how much redistribution is accomplished via taxes vs. via transfers, it’s better to use only market (pretax and pretransfer) income in calculating the effective tax rates. Doing that with the Tax Foundation data yields the following effective rates:

    Q1: 31.6% (bottom quintile)
    Q2: 28.1%
    Q3: 28.3%
    Q4: 28.9%
    Q5: 30.1%


  8. Social Security and Medicare taxes play a BIG ROLE. First, it is the majority of the federal tax that most folks pay (basically, anyone at the 50+% mark or below).

    Second, it is highly regressive — flat tax with a low cap. If person A makes a cool million next year, they’ll be throwing in about $10k to the SS/MC systems. If I make $110,000, it’s the same tax bill. So, the top $900,000 isn’t taxed, for person A.

    This offsets much the progressive nature of our income tax, especially in bringing the high end tax bills way down.

    That’s why Republicans only talk income taxes, it’s the only one that sticks it to the wealthy. Poor taxes are great!

  9. This recent Tax Policy Center ( report provides more data on income tax burden with the top quintile. If I read it right, roughly the top 10% of income earners do face a progressive income tax at the federal level; presumably this trend holds once we incorporate state, local, and payroll taxes as well.

    Inspired by this post, I wrote a follow-up on it here:

  10. I’ve been included in taxations for lengthier then I care to admit, both on the individual side (all my employed lifetime!!) and from a legal stand since passing the bar and following tax law. I’ve offered a lot of advice and righted a lot of wrongs, and I must say that what you’ve posted makes utter sense. Please persist in the good work – the more individuals know the better they’ll be equipped to cope with the tax man, and that’s what it’s all about.

  11. Social security houldnt be progressive. It was never intended to be a Welfare program, but basically a forced savings account. That’s why the cap is at 90k dollars… Ss is there to supplement your retirement, the uberrich don’t need the supplement, but they still pay (in most cases since they are self employed) roughly 12 grand a year, every year, as long as they are working, money that is essentially be redistributed to all the people who barely worked In their lifetimes and still collect social security.

    What’s gonna be a real drag though, is those of us under 35 will pay in for our entire lives and will likely not get anything from it, since it’s bankrupt, a ponzi scheme in it’s last days.

    And maybe it’s time to consider raising te age to receive ss benefits. When FDR introduced it, the age you received it was the age that the avg us citizen died. It was basically there to keep you from suffering if you happened to be one of the lucky folks that lived longer than you were supposed to.

    Now, ss income is used to live in gated communities in Florida, play golf and shuffleboard every day, etc.

  12. If you look at the bottom of Table 5 of the CBO report, you see that the taxable income on which the report is based is only $9.7 trillion. In that year, there was about $2 trillion of additional national income based on the BEA report. It is overwhelmingly likely that the additional economic income is to the benefit of those at the top end of the scale, particularly since the CBO includes cash non-taxable income to the poor but not non-cash income of the well-to-do. The report essentially runs off of the definition of taxable income of households and corporations. Thus, the total tax system may very well be regressive, taxing the rich a lower share of their economic income than the poor.

    Social insurance benefits should be means-tested. That would do a lot for the shuffle-board subsidy.

  13. Does this bar chart shown by Kenworthy take account of government payments to the less well-to do? Following his link, I went to the Tax Foundation webpage and found the following:

    “Overall, we find that America’s lowest-earning one-fifth of households received roughly $8.21 in government spending for each dollar of taxes paid in 2004. Households with middle-incomes received $1.30 per tax dollar, and America’s highest-earning households received $0.41. Government spending targeted at the lowest-earning 60 percent of U.S. households is larger than what they paid in federal, state and local taxes. In 2004, between $1.03 trillion and $1.53 trillion was redistributed downward from the two highest income quintiles to the three lowest income quintiles through government taxes and spending policy.

    “These findings suggest tax distributions alone do not tell Americans how much the nation’s fiscal system is helping or hurting low-income households. To answer that, we must look beyond tax burdens to government spending as well. Lawmakers who ignore the distribution of govern­ment spending risk making policy judgments based on an incorrect set of facts about the United States fiscal system.”

  14. I question this analysis. Federal Income taxes are such a large percentage of the tax burden on higher income folks that that factor alone would make overall taxes more progressive than this analysis suggests. For example, a high income person tends to pay a much higher percentage of his/her overall taxes in the form of FIT than any other form of tax. I would have to see the details to see what assumptions are being made by the authors, who appear to have a certain political agenda to further here.

  15. What slays me here is how the majority of Americans let the wealthy define the debate, when we have the votes to demand really progressive taxation. And we will try to avoid the ‘fair’ tax effects here, but it may be a good example. Also see L-curve site on Google.

  16. This chart is total non-sense and is designed to obscure to progressive nature of our tax code.

    First, Ken designed this chart to paper over what even his own curiously selected datasets show is plainly progressive. By displaying the data as “share” of income and “share” of taxes and by omiting meaningful labeling he effectively masks to the casual observer how much less the relatively less affluent actually pay in taxes as a proportion of their OWN income (which is WHAT we are talking about here). In other words, if the “poor” in this dataset average a 3.8% “share” of pre-tax income and pay 2.8% “share” of the total taxes the casual viewer is invited to only notice what appears to be a ~1.7% absolute difference on this chart between them and the very richest (and may be masked the fact that they’re moving in different directions) . However, the math doesn’t work this way (hence the reason he avoided real numbers) — you need to look at the relative differences for each quintiles tax vs income, not compare them to each other [that is, if this data were actually valid].

    Using his own poorly selected data, I’ve calculated the effective tax rates to be:
    Q1: 27.37%
    Q2: 32.08%
    Q3: 36.54%
    Q4: 41.76%
    Q5: 38.20% [Yes, this is reflected in his chart and it looks funny because it’s wrong.]

    Second, he did this by splicing the 2004 CBO pre-tax income data [table 1] for each quintile for the income side of his chart into the Tax Foundation’s total tax by quintile, but, whether he appreciates this or not, this data CANNOT be compared directly. The CBO and Tax Foundation group their quintiles using very different methodologies which means that they actually have different “individuals” in each and, in fact, actually different number of individuals (vs households) in each. The math simply doesn’t work and this is why you see such obviously screwy numbers for the rich (though it skews them all).

    Third, Ken doesn’t mention this in his post, but the Tax Foundation actually publishes this data as a percent of income, i.e., EFFECTIVE RATE, because he probably wants to use his own mathematically unsound and ideologically twisted “methods”.

    See this:
    Figure 3 (page 24) and Table 20 (page 48) for this. You can also see Figure 2 (Page 23) if you want to view a similar graph as what Ken presented that expresses this in “shares”, albeit with proper labeling, detailed explanations, etc.

    Lastly, Ken’s subsequent discussion on the comments here is totally non-sensical and bears little relationship to his chart and discussion about progressive taxation. You can’t pick and choose your datasets like this. To select the extensive datasets on taxation that the Tax Foundation has compiled, which mainly add in the sales and property tax component, but then expect the informed reader to ignore that said person is “spending” and getting taxed “more” as a result [esp. sales & property taxes] of non-market income (social security, welfare, food-stamps, etc) is quite ludicrous and will, in fact, tend to substantially increase the apparent taxation rates on the poor and lower middle class. His figures don’t tell actually tell us how much more the poor pay in these non-federal taxes on their hard earned income [the tax foundations indicates a roughly flat effective rate at the state and local level, roughly ~10%] nor do his supposed effective rates tell us the implied marginal rates the poor face on their actual labor (i.e., what incentive said person has to work harder, risk starting a business, and so on).

  17. Sorry, I meant “Lane”, not “Ken”.

    By way of comparison, the Tax Foundation estimates total effective tax rates (including local and state taxes) for 2004 to be:

    Q1: 13.0%
    Q2: 23.2%
    Q3: 28.2%
    Q4: 31.3%
    Q5: 34.5%

    Further, 2004 was likely one of the least progressive years a long time because of the Bush tax cuts and the fact that the market and the economy wasn’t doing so well in 2004. If the Bush tax cuts are allowed to expire (in whole or in part) and the economy doesn’t actually crash further, I would expect rates to be at least as progressive as they are in 2000 and probably more so (since the “rich” now have some extra taxes coming in around 2013]

    Here are the 2000 rates:
    Q1: 15.5%
    Q2: 26.3%
    Q3: 31.7%
    Q4: 35.5%
    Q5: 39.6%

    … all of this data can be reviewed in the PDF I cited above, which Lane apparently thinks is good enough to trust their tax data on, and you can actually see how they produced these estimates if you wanted to try to re-produce them [involves non-trivial statistical analysis and various assumptions].

    *Hint* *Hint*

  18. The share of total federal tax liabilities paid by any one income quintile is irrelevant. Taxes have been greatly reduced for the nation as a whole. Recent OMB figures show that in 2009, federal tax revenues accounted for about 14.8% of the GDP. This is a record post-WWII low.

    The real measure of progressivity of the tax system is the change in effective rates of taxation. According to the CBO, the top one percent of income earners a 31% effective rate in 2006, down from over 40% before the Bush tax cuts. Also, the income of the rich has risen much faster than that of the middle class which explains why they seem to shoulder a “disproportionate” share of the federal tax burden.

    Since the capital gains tax was reduced to only 15%, the very rich have experienced the greatest reduction in their effective federal tax rates;

    “The top 400 households paid 16.6 percent of their income in federal individual income taxes in 2007, down from 30 percent in 1995. This decline works out to a tax cut of $46 million per filer in 2007, or a total of $18 billion in tax cuts for these households per year…The decline in effective tax rates at the very top is due in large part to the capital gains tax cuts enacted in 1997 and 2003. The top marginal tax rate on capital gains is now 15 percent, less than half the top tax rate on wages and salaries. The top 400 taxpayers derived two-thirds of their income from capital gains and qualified dividends in 2007.”

    Though effective tax rates have tended to rise with income, the progressivity of the US tax system has been diminishing over time. This is especially true since the financialization of the US economy has shifted more and more income toward capital gains as the economy concentrates.

  19. “The share of total federal tax liabilities paid by any one income quintile is irrelevant.”

    Agreed — “shares” of taxes and income confuses this question and are generally not very helpful.

    “Taxes have been greatly reduced for the nation as a whole. Recent OMB figures show that in 2009, federal tax revenues accounted for about 14.8% of the GDP. This is a record post-WWII low.”

    True, tax rates have dropped for everyone, but incomes have also increased meaning that people pay more in terms of effective tax rates. It’s a serious exaggeration to claim that tax revenues as a share of GDP have fallen “greatly” when the variance has been a percent or two on both sides. The average receipts as a share of GDP was about 18% and the generally trend has not substantially changed. In fact, the average before the recession was actually slightly above post-WWII averages despite the Bush tax cuts.

    See the tables for yourself from the left-leaning Urban & Brookings Institute:

    “According to the CBO, the top one percent of income earners a 31% effective rate in 2006, down from over 40% before the Bush tax cuts.”Sounds like a cherry-picked number, but citation needed.

    “Since the capital gains tax was reduced to only 15%, the very rich have experienced the greatest reduction in their effective federal tax rates”The top 400 households are extreme outliers and these don’t have much impact on the top 1%, or 0.1% or 0.001%…

    If you want to implement a massive billionaires tax on Bill Gates and company, be my guest, but don’t assume that this reflects the majority of other tax payers.

    “Though effective tax rates have tended to rise with income, the progressivity of the US tax system has been diminishing over time.”

    Wrong. The US tax code has generally gotten increasingly more progressive and the rich getting richer generally would generally increase their effective tax rates. Although capital gains rates are essentially flat and have fallen, most of the rich get a minority of their “income” through capital gains and most of those that do are not rich like Bill Gates.

    Further, it’s a real mistake to confuse capital appreciation with income as you think of it. Someone that invested say, 100K in 1995 that did no better than the rate of inflation today would have an apparent 44% “gain” today even though they’re no richer (in fact, poorer in relative terms because of it). Add to this substantial risk premiums necessary so that people can break even on riskier investments (e.g., stocks). In other words, a lot of the “gains” are illusory and high capital gains rates can make investment non-worthwhile.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s