Some tax stats!

That double as fun facts?

If you haven’t already, you’ll be hearing a lot about taxes over the next two years. In this stack, I’ll give some policy context for that discussion, both why we are talking about taxes so much and some statistics about the US tax system.

Why now?

In 2017, Congress passed PL. 115-97, which is a nameless piece of legislation known familiarly as the Trump Tax Cuts or the Tax Cuts and Jobs Act. The act had two main provisions, a corporate tax cut and reform and a household tax cut and reform. The corporate cuts were permanent, but all of the household provisions expired at the end of 2025. The way legislation is scored by the Congressional Budget Office is through a ten-year window, in this case, 2018-2028. In order to reduce the cost of legislation over that window, a lot of bills, including this one, “turn on” and then “turn off” again. The hope, and it was certainly the hope here, is that the household cuts would be extended before they expire so a permanent policy is achieved but through a roundabout legislative way. You know with lots of uncertainty for the American tax payer in the meantime.

Unfortunately, there’s precedent here. The current era of tax cuts began with big cuts in 2001 and 2003 that, like the 2017 cuts, had almost all its provisions expire. In 2012, Congress passed another tax cut that extended (hard to say exactly but probably around) 80% of the 2001 and 2003 expiring provisions.

The question now is: are they going to do that again? Republicans certainly want to. I testified in front of the House Ways and Means committee last month as the lone Democrat witness next to four Republican witnesses as the majority members of the committee talked about how wonderful, successful, important, amazing, whatever this tax cut was and why they need to save it. I mainly talked about how it was a waste of money compared to other priorities. Didn’t go over well!

Last week, the Congressional Budget Office chimed in: extending the 2017 cuts will costs over $4 trillion. It would be the most expensive piece of legislation in history by at least a trillion.

Save for later

Are tax cuts good? Most of the time they are proposed as being beneficial, if not necessary, for the economy. My take is that they do help the economy some, but not as much as proponents claim, and at enormous expense. For being as expensive as they are, they should do a lot more. If you think of them as an investment, they are high cost with minimal return. But more on that later, I’m working on a paper about it. In the meantime, let’s put the US tax system in context with charts and charts and charts.

Charts!!!!

The IRS has publicly available data through its statistics branch that publishes summary statistics about the US tax system. There are thousands of tables to choose from, if you know where to look. Advocacy and research groups in this space, like the Tax Foundation, publish compilations of the IRS data in a more accessible format. But, however you access it, the data comes from tax returns.

So, with the universe of all tax filings, what do we know?

Well, we know their adjusted gross income (AGI)—this is the income that is 1) reported to the IRS and 2) subject to income taxes. So we don’t see income that is underreported (think sole proprietors) or income that’s not subject to federal income taxes (like a lot of Social Security benefits). Here’s the AGI for different cutoffs of the AGI income distribution—basically how much you have to be earning to be in the top 0.1%, 1%, etc. I show the earliest and latest year of data we have. (There are some technical reasons why data before 2001 isn’t comparable. But 2001 is a great year! That’s the year the first tax cut of this year was passed, so it gives a sense of the pre-period, or baseline, for all that followed.)

The top 0.1% cutoff went from $1.3 million to $3.8 million, while the bottom half cutoff went from $31k to $47k. Incomes at the very top are skyrocketing. Keep in mind this is the minimum threshold, and doesn’t say anything about the income above it.

I think this chart is really handy for thinking about incomes and inequality if you do some rounding. You make a quarter of a million dollars, you’re in the top 5%. You make $100k, top 25%; $50k, top 50%. And this is household income, not individual income.

Next, let’s add up all of the income reported to the IRS and parcel it out into these percentile groups.

Between 2001 and 2021, the top 5% of households (inclusive of the top 0.1%, rest of 1%, and 1-5%) took home a larger share of total income, the 5-10% took home a stable share, and the bottom 90% took home less. I can do some fast math for you: the top 1% earns 25% of all income. The top 10% earns more than half of all income. The bottom 50% now just takes home 10% of total taxable income.

In the 20 years depicted here, Congress has cut taxes four times. How did that affect these groups? Our tax schedule works by applying a marginal rate to taxes within an income range. The first $11,600 is taxed at 10% rate, but the next $36,000 is taxed at a 12% rate, and so on up the income level. Congress has cut taxes by adjusted the schedule as well as changing credits, deductions, and other policy. The average rate is not a policy set by Congress, so much as the result of those rates, filing behaviro, and incomes.

You can see in the next figure our progressive tax system at work, it shows the average tax rate for each income group. Every group pays less now on average than they did in 2001. They weren’t all cut by the same amount.

So now let’s look at the difference between those two lines, the change in the average tax rate for each income group over a 20-year period. The biggest drop was for the top 0.1%, 2.43 lower rate, the smallest was for the top 1-5%, 0.47, and the bottom 50% dropped 1.57 points.

So unequivocally and categorically: the group that has benefited the most by tax cuts in terms of their tax rate is the very top 0.1%. It’s about 150,000 households.

But here’s the part that can get confusing. Even though they are paying a lower average rate, their incomes are climbing enough that the share of taxes that they are paying is climbing. Rather than sum up all income and parcel it out, this next figure sums up all income taxes and parcels it out.

The top’s share of income taxes is also soaring, even as their average rate is dropping the most, because their income is increasing the most.

BIG REMINDER HERE: this is just federal income taxes, not all taxes. For Social Security and Medicare, which collects two-thirds in total of what income taxes collect, this share-of-taxes-paid figure would look very different, because payroll taxes are regressive and have a tax cap. So the top 10% contribution would be relatively little compared to the bottom 50.

Makers and Takers

That is probably enough charts for one day. But I want to include a note on makers and takers. In the 2012 campaign, Mitt Romney said that 47% of Americans pay nothing in taxes, that they are the takers in our economy/society/country. He was pushing out a myth that many have before and since: that there’s this mass of permanent lazy people who pay nothing and get benefits in return.

You can reference the figure above to see yes, in fact, the bottom 50% does pay income taxes, or reference the reminder above to know yes, in fact, the bottom 50% does pay a large share of payroll taxes.

A researcher at NYU took aim at this 47% notion and did a study where she looked at households over a 40-year period and their tax liabilities and transfer receipt. About 70% of us will have a zero tax year at some point in our non-elderly years. Income in the US is highly volatile and changes a lot over someone’s life (think about how much you earned at 20). She has a great explainer of her findings.

Final note, most people have a really wild sense of public program beneficiaries and costs. The bottom 50% comprises 75 million households. Food stamps, or SNAP, goes to 22 million households. Housing vouchers go to 2 million. About 7.5 million get Supplemental Security Income. Less than half a million households get TANF. The majority of the bottom 50% is not on a public transfer program.

(Health is a different story, there are about 76 million people on Medicaid, but you’ve got to do like for like here. The government is buying their health insurance but it is undoubtedly subsidizing yours. The value of employer provided health insurance is not subject to the income tax, this benefit flows to the top of the income distribution and at over 1% of GDP, is the largest tax break we have.)

The same thing goes for costs. If you added up spending on means tested programs—Supplemental Security Income for the elderly and disabled, Pell Grants for college students, school meals for K-12 students, food stamps for poor families, Temporary Assistance for Needy Families block grants, and the federal components of Medicaid and the Children’s Health Insurance Program—their combined spending is still less than Medicare.

This is probably why many people think the way to get control of the government’s debt is to cut these programs. If you think 75 million Americans don’t pay taxes but instead get 3-5 different public benefits, then of course spending cuts for the poor is the way to go. But the starting point is wrong, so the solution is too.