The Numbers
The data for the family income analysis and break-down of state and local taxes comes from The District of Columbia’s Office of the Chief Financial Officer. Information on U.S. Federal taxes and income, population, and GDP come respectively from the Internal Revenue Service, the U.S. Census Bureau, and the U.S. Department of Commerce Bureau of Economic Analysis. Comparative country data comes from the Organisation for Economic Co-operation and Development.
For the D.C. Office of the CFO reports in particular, and for much of this information in general, there are dangers in comparing across years. If you read the CFO reports it becomes clear that their methodology changes from year-to-year. Typically, the modifications are minor; however, in some cases there are significant discrepancies as a result in improvement in methodology. For example, the calculation of real property tax for the $25,000 household changed substantially between the 2004 and 2005 reports. In such a case, I recalculated the older data using a straight-line depreciation model from the newer data.
An even more complex problem arose in determining Federal income tax at specific income points. The Federal tax data is reported in bands (e.g., income from $75,000 to under $100,000; or income from $100,000 to under $200,000). As you can see from the example, the bands are inconveniently not of the same size. To handle this I calculate the midpoints of the bands and find the midpoints which surround the target income. So, to calculate the estimated tax at $50,000 income, I start by saying that $44,999.50 and $62,499.50 are the midpoints of the brackets that surround $50,000. Fifty thousand is then $5000.50 from the lower midpoint, which in turn is 28.5743% of the distance between the two midpoints (62,499.5 – 44,999.5 = 17,500; 5,000.5/17,500 = .2857428…). The difference between the average tax for the two brackets is then multiplied by that percentage, and the result added to the lower of the average taxes. Unfortunately, even this starting data is not banded for families of four with two dependants. My best estimation was to use the data for married persons filing jointly. This is a better starting point than using all data, or head-of-household, or any other category. However, it aggregates the taxes regardless of number of dependants — so it will include those families with one child, as well as those with six. If someone has a better source of data based on actual number of dependants, I’d be happy to use it and recalculate. Last but not least, the $150,000 target income falls in the $100,000 to less than $200,000 bracket. For all years, that bracket crosses the AMT threshold, which means the taxes expressed in the aggregate data are not directly comparable. For that reason, rather than using the above described methodology, the MicroSoft Excel Forecast function was used.
Social security calculation was fairly straightforward. Medicare is a flat 1.45% of income. Social security for the entirety of the period covered is 6.2% of income; however, the cap to which the tax is payed changes every year. Since the D.C. CFO report provides the breakdown between the two incomes in the household, we are able to directly calculate the exact tax.
On page 2, for calculations of the rich, the 1% is based on all tax returns for the year, starting from the highest AGI and working downward. In every year covered, around 1% of the population reported zero or negative AGI, with an average of a %52,000 loss. It is my belief that the vast majority of that lossful group is small business owners, and the remainder are people forced to move and suffering a capital loss from sale of a home, or very low income individuals (part time self-employed, or on unemployment benefits) that are also paying healthcare or some other deductible expense. But that’s just a belief. It’s possible that some people in this category are the super rich. If anyone has a credible source of information on who makes up those with negative AGI, I’m happy to adjust the calculations throughout this blog entry.
For the global comparisons, there are several sources for GDP and comparison of taxes across countries. I have selected the OECD database primarily because the total tax burden numbers for the U.S. they have calculated are within 0.1% of the numbers I have calculated. Given the difficulty in finding and collating this data, that degree of accuracy has me convinced they have done their homework. While other organisations (e.g., the IMF) may still have done their homework and may just be using diffeent economic definitions of GDP, the numbers from the OECD seem more directly comparable.
Last but not least, three items:
- Please, leave a comment — I hope this will spur some interesting conversation;
- If you uncover any errors, please email, tweet, or otherwise contact me, and I’ll do my best to correct them; and,
- If you want a copy of the excel sheets and the original data, register with the blog and send me an email, and I’ll be happy to send you a copy.
Chris,
This was a great post, informative and interesting. I have a few thoughts that perhaps you or someone else could comment on:
First, regarding the taxes paid by the wealthy (i.e. “top 1%”), many people combat the notion that the wealthy are being disproportionately taxed by invoking the law of diminishing (marginal) utility. That is, while the wealthy may be paying a considerably greater percent of their incomes in taxes than the less affluent, they, the wealthy, still have a ton of money laying around and the taxes paid, despite being numerically larger, are worth about the same when adjusted for utility. Inherently, this makes some sense to me in that I may more desperately need that X percent of my income if I make $20,000 per year rather than $200,000 or $2,000,000. However, a dollar is always worth a dollar and any concept of utility seems too subjective to be helpful on this scale. What are your thoughts on this?
Second, you made an excellent point on the costs associated with complying with our über-complex tax code. The cost alone is enough to justify considering alternate systems (flat tax, fair tax, etc.). But, how do we, the US, relate to comparable countries in regards to tax complexity? A quick skim of the World Bank Ease of Doing Business Rankings is informative for some glimpse at corporate taxes, however it was relatively unhelpful in yielding information on personal tax complexity. I would be especially interested to see which tax systems have stimulated the most growth.
Third, I am curious about the connection between our total tax burden and the inference that we have the ability to tax more yet still grow our GDP per capita. It is obvious that other countries are able to tax more and maintain a higher GDP per capita than us (or, as your rightly pointed out, raise their GPD per capita faster than us despite higher taxes), but I wonder if suggesting that ‘because other countries have done it, we can too’ may be a too simple an answer. I am certainly no expert in macroeconomics, but some of the countries that managed this feat (like Luxembourg, Switzerland and the Netherlands) are working with a significantly smaller populations, which, I imagine, could be more readily affected by an influx of businesses than our much larger country – in effect growing GDP despite the “tax burden.” So, the question is how much can we learn from countries like Luxembourg and where do those analogies lose utility?
Anyway, thanks again for the great article!
Best,
Micah
Hey Micah,
Thanks for the great reply and the great questions! I’ll come back to the first and third a little later, but I wanted to just give a quick answer to your second question.
The bottom line is that the U.S. doesn’t have any peers of comparable tax complexity. The closest is the U.K., at 11,500-ish pages is closest. But even there, that’s the entirety of their tax code. The over 16,000 pages I quotes for the U.S. was income tax only, and Federal only. After the U.K., other countries drop off quickly. Canada’s income tax act is over 2,000 pages — but it’s written in two languages at that length! The French weigh in at 1,900 pages. After that, it falls off steeply. Of the countries in which I’ve lived, Germany is pretty big at 141 pages, but that’s for the entire Abgabenordnung (or fiscal code, not just the taxes). The Czech Republic, Russia, and Singapore all have flat(-ish) tax structures these days. Really, it’s almost impossible to compare.
As somewhat of an aside, I suspect part of our tax complexity is due to our age. In Europe (and even in the U.S.), people think of the U.S as a young country. However, the reality is that we’re one of the oldest countries in the world. There are only 3 countries in Africa that predate the 20th Century, and none that predate the 19th. In the Americas, a few date from the 19th Century, but we’re the only one to date from the 18th. In Asia, Thailand dates to 1776, but it’s the only 18th Century country, and the vast majority are 20th Century. In Europe, San Marino dates to 1600, the U.K. dates to 1707, and … that’s it. The U.S. is the fourth oldest country in the world. That gave us a lot more time than anyone else to mess up our tax code.
Once again, thanks for the reply, and I’ll get back to your other two questions as soon as I have a few moments.
Hi Micah,
Coming back to your other questions, regarding the first, I’m not sure I completely buy into the marginal utility argument, at least not from that direction. It depends on what one means by “worth”. Money is an abstraction of saved societal free-time, and the rich (who by definition have acquired large quantities) are capable of doing large things (witness Bill Gates philanthropic efforts, or Sir Richard Brandson’s X Prize). One can argue that these have greater worth to society and are a natural outcome of accumulation of wealth. I would say those dollars are worth more to society, not less. That said, they are certainly worth less to the individual (i.e., you’re correct, those saved dollars are more meaningful to the poor than they are to the rich).
The question ultimately comes down to what a “fair” tax is. As I mentioned, the top 1/2 of Americans pay 98% of the Federal income tax. Digging further into this, the top 10% of Americans pay 45.1% of total taxes. This is higher than any other OECD country (for Germany, the number is 31%; for France the number is 28%). This is true even though France and Germany have much higher marginal tax rates because they also have much higher consumption taxes — and consumption taxes (e.g., sales tax) are regressive.
Regarding your third question, yes, it’s hard if not impossible to compare the U.S. to any other country. There are serious issues with scale, which is why I selected the per capita number. Two other ways we could have compared would have been to look at countries with similar populations, and to look at countries with similar total GDPs. Looking at pure GDP, there’s really no one comparable. At $13.8 trillion in 2009, the U.S. was slightly smaller than the entire E.U., but slightly larger than the 17 country Euro zone, 35% bigger than China, and more than triple the next closest economy (Japan). So, I suppose we could compare to China, but I think even more than in the per capita measures, country specific items other than taxation policy would drive the conclusion. Going back to similar population, it’s slightly better than similar GDP, but not much. In 2009 there were 5 other countries which had the same order of magnitude population as the U.S. (which is to say, in the hundreds of millions): Indonesia, Brazil, the Russian Federation, Japan, and Mexico. I would argue that the only one of those worth comparing is Japan. As it happens, the data isn’t even available for Indonesia, Brazil, or Russia. I’ve put together another chart for Mexico, Japan, and the U.S. here: https://www.cwrichardson.com/images/TaxVPopChart.png and I think you’ll find it no more illuminating than the per capita data.
But again, you’re correct. It takes miles to turn a tanker. Thinking that the sort of economic change that caused the success of Switzerland or Ireland could be effected in the U.S. would indeed be naive. I’ll write a separate post on what I think the policy implications of this data should be; for the moment, my only point was that changing tax policy is not going to have a meaningful economic impact.
Cheers,
Chris
Chris,
For the sake of simplicity, I’m going to respond to both of your replies here.
First, I have to admit that I was one of those people that assumed the US was a fairly newcomer on the country-founding scene, at least in comparison to Europe and parts of Asia as I’m fairly familiar with Sub-Saharan Africa and Latin America. I know that I visited a few bars in Prague that are older than the US, so to learn that we are the fourth oldest country in the world came as a shock.
Being an attorney by training, but not by trade, I can say that our legal system has generated similar levels of relative complexity in all areas of the law when compared to our Southern neighbors. When I’m working on projects in Latin America, it always amazes me how easy it is to maneuver in a civil law setting – most things are fairly well laid out in the code and it is reasonably static by comparison. Though, it has been interesting to watch the evolution of legal systems throughout Latin America as they slowly shift to include more of what we in the States would consider to be a common law approach. Of course, in many respects we’re moving more and more towards a codified legal system, so perhaps we’ll meet in the middle. All of this is to say that maybe “younger” countries will catch up to us in tax complexity as they age or maybe we should consider pursuing the sweeping reforms that many countries enact (seemingly without much trepidation) on a fairly regular basis.
Second, I completely agree with your comments on the marginal utility argument. I particularly like your view on the societal benefits the wealthy can generate. While some may debate whether those benefits are a natural outcome of wealth accumulation, history is filled with corporate tycoons turned philanthropists and I suspect this trend will continue.
Regarding the relative percent of taxes paid by the wealthy, I wonder what the effects of wealth distribution, or the gap between the rich and the average, have on the percent of the population paying the majority of the total taxes by country. That is, does it take a smaller percentage of the population to pay the majority of the taxes in the US because that percent of the population is so much wealthier than the average taxpayer? So, setting aside for a moment the debate on whether the wealthy are paying enough or too much in taxes, can some of the difference between the US and Germany and France, in terms of what percent of taxpayers are paying the majority of the taxes, be accounted for by disparities in the number or magnitude of the wealthy in each country? If so, how significant is this effect?
Third, thank you for your brief analysis of alternate means of comparing the US to other countries. Admittedly, I find all of the comparisons to be fairly unsatisfying though, sadly, I can’t recall enough math or economics to suggest additional means of comparison.
I’m looking forward to your policy implications/recommendations write up. I don’t know where you’re finding the time to work on all of this, but I am thoroughly enjoying it.
Cheers,
Micah
Good stuff! Maybe you should be in politics!
Great write up!
I’m skeptical of the state and local tax numbers. My CA effective tax rate is much less than my Federal effective taxe rate and CA very high state taxes.
I never understood the argument for a flat tax rate vs just eliminating exemptions. Marginal tax rates are pretty simple. People would understand the tax system better if they computed their tax from the marginal tax rate table instead of the precomputed tax tables. (IIRC this is how it was on the original federal tax form.)
I’m not sure if the rich are paying their fair share, at least not rich investors. 15% on long term capital gains is a steal.
Hey David,
Thanks! I’m definitely glad you enjoyed it.
It’s true, I would imagine your effective tax rate in CA is much lower than your effective Federal tax rate (see the first chart on page 1). However, your local and state taxes are still almost certainly higher than your Federal taxes (particularly in CA). When you add your property tax, auto tax, your CA SDI, and most significantly your sales tax, I suspect you would find those add to more than your Federal tax. That said, it’s possible they may not, as the analysis here was for hypothetical families of four, not for any specific individual. Also, the analysis was based on the largest city in each state, and living in LA could be very different than living in San Bernardino.
Regarding the rich paying their fair share, see my reply to Micah — maybe, maybe not. But as to tax policy, I agree. I’m not advocating a flat tax (even if we wanted one, we probably couldn’t implement one, because of state and municipal differences — I may have a flat tax in the Czech Republic, but that’s in part because I don’t pay separate taxes to the state of Bohemia or the city of Prague).
Now that I have excavated all this data, I do intend to come forward with some proper suggestions. In the meantime, I just wanted to point out some immediate things that jumped out of the data: that we have a preposterously complex tax system, given that for the middle class we’re effectively extracting a flat tax; and the rich are paying more than most people probably think they are.
With respect to the specific point of the 15% long term capital gains, I’m not sure. I don’t know if I can extract from anywhere the distribution of sources of income for the wealthy. But capital gains would imply they’ve sold some capital. I imagine that the majority of their income in most years comes from interest and dividends, which would be taxed as income. And in fact, capital gains and capital gains taxes are part of AGI, so both are included (though buried) in the above analysis.
Thanks again for reading and commenting!
Chris