To get us into the mood of looking at this, and as a set-up for starting the analysis, ask yourself this question: Am I in the top 1% of income earners?
Unless your Adusted Gross Income (AGI) for 2009 was $343,927 or more, you’re not. Here’s a table I put together from the latest IRS data that shows the breakdown from 2001-2009.
And here’s a graph that shows the same information.
This graph, however, gets so distorted by the incredibly high incomes of the 0.1% that it makes the rest of the data hard to read. Taking out the 0.1% numbers compresses the Y axis and makes the remaining 99.9% much easier to see.
One thing that jumps out in both charts is that most of the downturn in income in 2007-2009 came at the very top ends of income. At the top 50%, 25%, and even 10% levels, 2009 incomes were essentially the same as in 2007. At the top 5% and above, however, we see a downturn, with the downturn increasing as the incomes rise. Without data to refer to it is impossible to be certain, but this would seem to correlate very strongly to a loss in capital gains income as the stock market crashed. If that were true, as a result we would see a corresponding rise in the average tax rate paid by the top tiers in 2008-2009 (Capital gains tax rates are, on average, lower than the rates paid in normal income tax.).
The table here is also from the same IRS dataset. It confirms this relationship.
The rates for the top 3% bracket and above are all higher, the rates for the 4% and 5% bracket are basically flat, and the rates for the 10%, 25%, 50%, and bottom 50% brackets are all lower. These lower tax rates are as a result of stimulus measures and would have had a similar, but smaller effect on the upper brackets as well had they not had such a high portion of their income coming from capital gains (and potentially tax-free municipal bonds or similar instruments). Here’s the same data in graph form.
An important point here is that the losses in the stock market, unlike losses in other income due to a cutback in hours or a lowering of one’s salary, can be deducted in following years. This means that during a year with a loss, no capital gains income is recorded and other forms of income become a higher portion of AGI, which raises the average tax rate. When stock markets recover, the losses from the previous years can be offset, which brings the cash back into the investor’s hands, but again is not recorded as taxable income. This keeps the AGI at the higher level on the remaining income, but vastly reduces it when compared to cash flow, or actual income.
So it’s clear that the very wealthy – the top 1% — are paying a higher tax rate than everyone else, but they also get some advantages from the tax code that allow them to write off losses that would not be available to a person not able to invest heavily in the stock market. I’m not here to argue whether this portion of the tax code is correct or not, but is important to understand that much of the income fluctuation in the top brackets is not coming from lost wages, but from lost investments, and that our tax code allows those losses to be recovered in future years.
We often hear the claim that the top 1% pay a greater share of the tax burden than the bottom 90% combined. That’s only correct when talking about income taxes, however, and the payments made for Social Security and Medicare are often ignored because they make the statement much less strong. We can do some easy calculations against this IRS data to see what the impact would be, so let’s take a look at the results.
In 2009, the Social Security tax rate was 6.2% on the first $106,800 of income. No Social Security tax is paid on income above this level. Medicare is taxed at 1.45% for all wages. In order to do this analysis, some assumptions need to be made about what portion of income at the top tiers comes from non-wage sources. I tried to find data on this and was unsuccessful, so I selected high numbers – numbers that would tilt the analysis in the direction of showing the upper tiers paying more taxes – in my analysis. The table here shows what I did.
Notes:
[1] This analysis estimates Social Security income by using data from the CBO here and looking at the 50% employee contribution. Because the purpose here is to show how the addition of Social Security and Medicare payments greatly shifts the tax burden to the lower income brackets, I am not overly concerned about absolute accuracy. It is important to note that Social Security is capped, which means that you never have to pay for it for incomes over a certain level ($106,300 in 2009). The result is that for 2009, everyone in the 10% and higher income bracket paid only up to that amount. It also means that any shortfalls for this simulation must therefore come from the brackets below 10% and would shift the tax burden in that direction.
As a graph it looks like this – Note the comparison to straight federal income tax payment shares. It shows very clearly that the additional of FICA to the analysis makes a huge difference and absolutely should not be ignored when discussing overall tax burdens. With FICA included, the top 1% doesn’t even cover as big of a share as their neighbors in the 1.1-2.0% bracket, although this bracket’s share shrinks as well. Once into the lower 98% of earners, adding FICA greatly increased their share of overall taxation and changes the graph dramatically.
It is obvious that taxpayers in the higher income brackets pay more taxes, and even pay a higher percentage of their gross income in taxes, when compared to lower tax brackets. What is absent in the discussion is a reasonable look at what percentage of disposable gross income is paid in taxes. Without this calculation, it is not possible to evaluate and discuss the impact different taxes have on different income brackets.
Disposable gross income could be calculated in many different ways. A simple, straightforward, and reasonable baseline to begin the discussion would be at the poverty level. US Census Bureau poverty levels for 2009 and other years can be found here. Poverty levels are calculated for size and age of households, but because of the limitations in the data from the IRS, we will have to select one dollar amount to use for the analysis. The average household size in the USA right now is around 3 people. It was smaller in the past, but economic constraints are leading to larger households. Because of this, for this analysis I will use a three-person household with one child as the poverty baseline. That leads us to $17,268 as our poverty threshold.
The way I calculated the tax burden as a proportion of disposable income is as follows: Federal Taxes Paid (With and without FICA)/(Adjusted Gross Income – Federal Taxes Paid – Poverty Threshold. I found the average (As opposed to split levels) incomes for each bracket by taking the total taxes paid and dividing it by the total number of filers in each bracket. For those of you who were happy that you squeaked into the 1% bracket by crossing the $343,000 level, take a deep breath: The average income for members of the 1% bracket is $960,000. Here is the table.
And as a graph.
Note that the bottom 50% of Americans are showing a tax burden now that actually exceeds their disposable income. In reality, most of these people are somewhere between doing OK and scraping by, but just as the ultra-rich at the 1% and higher end of the spectrum have their peaks skewed due to people like Warren Buffet, the lower brackets will tend to balloon outward as well. This is why we have tax credits, welfare or food stamp programs, and other systems to support these people. My belief is that if we analyzed the lower brackets we would see a more steady and less abrupt progression, with people in the lowest quartile banging up against the 100% tax burden calculation in this analysis. Similarly, a more detailed analysis of the very rich, for example at 0.1% increments after 1%, would show a much more even distribution with taxes rising from 1.0% to 0.5% on a slope more similar to what is seen from 5% through 1.1%, and then falling as it approached the 0.1% level.
Please also note that the peak in this calculated tax burden is right at the bottom of the top quartile, or just as one moves out of the middle class and into the upper middle class. Much of this is a function of two things: The cap on Social Security taxes, and the assumptions I made about proportions for wages to income.
The instinct here may be to point to the peak at 1% and use it to argue that the rich are paying their fair share or even more than they should be, but when considering for how much disposable income remains after payment it is clear that there is no financial duress at those levels.
I’m very open to comments and criticism on this. If anyone wants the Excel files for the data, leave a comment and I’ll e-mail it to you.
Recent Comments