Not quite the same as this related question but similar.

According to Wikipedia, the origin of the "Buffett Rule" oft-quoted by former-President Obama is that:

rich people, like himself, could pay less in federal taxes, as a portion of income, than the middle class

According to the related question, it is true that Buffett himself did indeed pay a lower tax rate than his secretary in the year quoted (his secretary being the proxy in the particular example of "middle class"). However, in a discussion I was having over on Politics SE, I was linked to this article, which shows (or appears to show) that, in this respect, the relationship between Buffett and his secretary was an outlier in terms of the relative tax burden of the rich and middle-class, and, in general, the rich do indeed pay the most tax as percentage of income, and furthermore, among the rich, the top 1% pay the most-most tax.

What I want to know is, how much factual basis does the "Buffett Rule" have? Is it actually true that the rich pay a lower tax rate than the middle class, as Buffett stated, or is Buffett and his secretary simply an outlier of a system which is otherwise working properly?

  • Fig 3 in the document cited indicates around 10% positive (not negative) income tax for the lowest bracket. They actually seem to include various benefits ("in-kind benefits from social insurance and other government assistance programs") in the "average" case (fig 6) to claim the tax rate is negative. Not everyone may claim those. – Fizz Oct 1 '18 at 19:44
  • 1
    If I was to take a completely uneducated wild guess, I would say that the -9.2% figure probably mostly derives from public assistance programs, which the lowest income bracket take advantage of but do not proportionally pay into. – Ertai87 Oct 1 '18 at 19:51
  • 1
    One thing to remember is the gap in how much more the rich make then the middle class so even if they pay a smaller percentage the amount they pay is a lot larger. For example someone making 150k a year paying 15% will pay more overall then someone making 100k a year paying 20%. – Joe W Oct 1 '18 at 21:35
  • You need to decide which taxes to consider. With FICA taxes, you also need to consider the "employer portion" - which is part of the pay from the employer's point of view, but which is not subject to the employee's individual income tax. – David Thornley Oct 2 '18 at 17:55
  • It is worth noting that the tax rate that should matter for individual decision making isn't the headline rate of tax they pay but the effective marginal rate when all taxes and benefits are considered (which takes into account the rate at which subsidies are withdrawn when marginal income increases). I don't know the US situation, but the UK has some barking mad marginal rates at both ends of the distribution (eg a £1 extra in income from £100k/yr adds about 20% temporarily to marginal tax meaning those earning £100k pay way more than those earning £150k). – matt_black Oct 11 '18 at 18:31

This is highly contentious depending which taxes (and benefits) are included. The wiki page only includes federal income tax offset by ("in-kind benefits from social insurance and other government assistance programs") from a CBO report fig 3 (nominal tax) vs fig 6 (offset by benefits). And they find a huge gap, with the lowest bracket having a negative effective federal tax after benefits are subtracted.

On the other hand, Piketty et al. aggregating multiple tax sources "federal, state, and local levies—including corporate, property, income, estate, sales, and payroll taxes" conclude that the gap is much smaller:

enter image description here

Also the category of "rich people" is pretty vague. The top 1% may be too broad for that. The effective tax rate actually goes down as we subdivide that:

enter image description here

From WaPo's Christopher Inghram (2015), who is critical of this state of affairs:

In other words, a person in the top 0.001 percent income bracket -- who would have an adjusted gross income of at least $62,000,000 -- pays the nearly same effective tax rate as somebody in the top 20 percent bracket who makes $85,000 in adjusted gross income.

That's not how federal income taxes were, at least originally, designed to work. The super-rich pay a relatively low rate for a variety of reasons. They benefit from a whole host of deductions -- like the mortgage on a yacht, for instance -- and other tax benefits that many people don't qualify for.

Chief among these is the lower tax rate on capital gains -- think investment income. That maxes out at about 24 percent when you factor in a Medicare surtax that applies to some investment income. But wages are taxed at a top rate of 39.6 percent. Since many of the super-rich get most of their earnings from investments, they disproportionately reap the benefits of that lower capital gains tax rate.

In the year this data was compiled, 2012, the top capital gains rate was lower still, at 15 percent.

That graph probably has changed somewhat post-2013 reform. For 2014:

The top 1 percent of taxpayers (AGI of $465,626 and above) paid the highest effective income tax rate, at 27.2 percent, 7.9 times the rate faced by the bottom 50 percent of taxpayers.

Taxpayers at the very top of the income distribution, the top 0.1 percent (with AGIs over $2.14 million), paid an even higher average tax rate, of 27.7 percent.

And taxes were slashed by Trump in 2017 with "pass-through" companies supposedly mostly benefiting the rich.

Individuals in the bottom 80 percent earn virtually no pass-through income. Moreover, those with higher incomes tend to receive a much greater share of their income from business compared to those with lower incomes, as the top 1 percent only earn about 11 percent of wage and salary income. Thus, any reductions cuts in the tax rate on pass-through businesses would largely benefit high-income taxpayers.

enter image description here

According to a Bloomberg article:

The new law lowers the top individual rate to 37 percent from 39.6 percent. With the new 20 percent deduction, pass-through owners taxed at the top rate can now get their rates as low as 29.6 percent.

Critics say this creates an incentive for top earners to recast themselves as independent contractors and funnel wages taxed at ordinary rates through a pass-through entity. [...]

The law starts phasing out the pass-through deduction once the net income of an owner in one of those professional fields hits $157,500, or $315,000 for joint filers. Once their income hits $207,500 -- $415,000 for joint filers -- the deduction disappears altogether.

So the calculations would change again. I haven't found an updated version considering this. In that March 2018 Bloomberg piece, various experts were chiming in how easy it would be for the rich to [ab]use the new law.

[Proposed] regulations accompanying the new law have only been recently announced as of Aug, 2018, which probably explains some uncertainty of their impact.

So it's difficult to answer this question in general, without fixing a year or deciding who is the "rich".

  • 5
    A corollary that appears based on that diagram: When the US is at war, the top 1% disproportionally pay the costs of the war. See the spikes around 1938-1945 for WWII, mid-1950s for Korean War through to early '70s for Vietnam (except for roughly the decade 1958-1968, not sure why), mid-90s for Desert Storm, and early-00s for the Iraq/Afghanistan wars. Not sure if the spikes are actually related to war spending, but it seems correlated. Not at all related to my question, just something I found interesting. – Ertai87 Oct 1 '18 at 20:09
  • In summary, would it be correct to say the following, retrodated to 2012: The top 1% are paying the most amount of tax as percentage of income, but from there the effective tax rate decreases until someone in the ultra-super-rich are paying a tax rate comparable to the middle class? According to Google, the population of the USA is roughly 325MM, so the top 0.001% would be about 3250 people, so when we say "the rich should pay their fair share", the people we are really truly upset about amounts to about 3250 people out of 325MM. – Ertai87 Oct 2 '18 at 14:03
  • 2
    @Ertai87: if the "middle class" is the top 20%, which is probably a bit thin for the middle. – Fizz Oct 2 '18 at 14:05
  • Fair enough. In which case, would it be accurate to say, retrodated to 2012, that the whole "Buffett Rule" issue was more or less nonsense? – Ertai87 Oct 2 '18 at 14:23
  • @Ertai87: More or less. It's only true in a certain geometric sense, but not literally, at least that year. – Fizz Oct 2 '18 at 14:26

In general, if we look at income by quintile, we see something like (from a Congressional Budget Office report):

Average federal tax rates, by income group, 2010

In this, we can see that the average federal tax rate paid by the top quintile is higher than that paid by the lower quintiles. And we can see that the rates for the top 1% are more than those of the remainders of the top 5%, top 10%, and top 20%. In general, the top 1% and the remainder of the top 5% pay a higher rate than the top 20% (including them).

Another issue is that Buffett's "income" is almost entirely investment income, which we calculate quite weirdly. With investment income, we tax what are called capital gains or dividends. But we already taxed the money that was invested. And we tax the nominal value of the gain rather than the real (inflation-adjusted) value. This is called inflationary gains. See here.

Someone like Warren Buffett will have a lot of investment income and very little salary income. Meanwhile, his secretary has only salary income. So he pays a lot of tax on inflationary gains and she pays very little. From the link, the rate he pays on his real investment gains is 29% even if his rate on the nominal gain is 20%.

It's worth noting that when the graph was made, the top tax rate on capital gains was 15% but the inflationary gains link calculates based on the current 20% top rate.

  • Exactly what are those tax rates? Federal income tax? FICA? Other Federal taxes? Are they derived from actual taxes paid or the set rates? – David Thornley Oct 2 '18 at 17:51

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .