The distribution of TCJA cuts, as well as the burden of financing them, by income group and race
The House and Senate both passed versions of the Tax Cuts and Jobs Act (TCJA) in recent weeks. Both versions of the bill, which must now be reconciled and voted on again, are made up mostly of large, hugely regressive tax cuts that give disproportionate benefits to big corporations and the wealthiest Americans. While the regressivity of these bills by income class has been well-documented by now, we’ve been asked by a number of people about the likely distribution of tax cuts called for by the TCJA across racial groups. A fully fleshed-out and precise estimate of this racial distribution would take lots of time and effort to calculate, but a decent rough estimate can be made pretty quickly if we’re willing to use some plausible proxy data.
However, it is also crucially important to note that congressional Republicans have not just passed versions of the TCJA in recent weeks, they have also passed a budget resolution calling for steep cuts to key programs, in large part because they want this money to finance their tax cuts. Assessing the impact of tax cuts while ignoring likely spending cuts would lead to a radical underestimate of the effect of coming fiscal policy changes on typical Americans’ livelihoods. Given this, we also examine the likely distribution of the burden of financing the TCJA with spending cuts by income class and race.
The Urban-Brookings Tax Policy Center (TPC) has provided estimates of what share of the tax cuts would go to different income groups. The Survey of Consumer Finances (SCF) provides data on the share of households in each of various income groupings that are headed by white, African American, or Hispanic householders. The SCF is uniquely useful here because it has clear income percentile rankings all the way up to the top 1 percent. Merging the TPC and SCF data in this way is not a pure apples-to-apples comparison. The TPC data is arranged by “tax units” while the SCF data is arranged by households (while the SCF calls their unit of analysis “families”, it is much closer to the “household” definition used by surveys like the Census). A tax unit can contain more than one household. But, all this said, there still should be substantial overlap between the two data measures, and the TPC data on tax units should provide a useful overview of the distribution of tax cuts across households.
The first row of Table 1 below shows the current distribution of income by income percentile class. The next four rows show the share of each income percentile class that is white, African American, Hispanic and “other” (in the coding of the SCF). Unsurprisingly, white families constitute the majority of every income percentile group. But it is striking (if not surprising) that their share is lowest in the bottom two-fifths of the income distribution and highest in the top 1 percent. Both African American and Hispanic shares are highest in the bottom two fifths and fall monotonically as one moves higher in income percentile rankings. The upshot of this for the distribution of policy changes is crystal-clear: regressive policy changes are likely to inflict the most damage (or provide the least benefit) to non-white families.
Racial composition of income groups and distribution of TCJA benefits and financing burden, by income percentiles
Shares of… | Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total or weighted average |
---|---|---|---|---|---|---|---|---|---|
Total income, 2017 | 4.4 | 8.7 | 14.2 | 20.7 | 14.2 | 9.8 | 12.9 | 15.3 | 100.0 |
White families, % of total | 59.3 | 68.7 | 74.7 | 79.6 | 83.8 | 85.7 | 89.5 | 92.7 | 73.6 |
African American families, % of total | 24.2 | 15.8 | 12.3 | 9.3 | 7.4 | 5.6 | 2.2 | 1.3 | 13.5 |
Hispanic families, % of total | 12.4 | 12.5 | 9.3 | 7.1 | 4.2 | 3.0 | 2.0 | 2.1 | 8.9 |
Families of other race/ethnicity, % of total | 4.1 | 3.1 | 3.7 | 4.0 | 4.6 | 5.8 | 6.3 | 3.9 | 4.0 |
Benefits of TCJA, House | 0.3 | 1.3 | 8.7 | 15.7 | 5.5 | 1.6 | 18.9 | 47.2 | 100.0 |
Benefits of TCJA, Senate | -1.0 | -0.3 | 2.9 | 7.3 | 7.2 | 5.7 | 15.7 | 62.6 | 100.0 |
Burden of financing with lump-sum taxes | 27.9 | 22.1 | 19.6 | 16.5 | 7.1 | 3.4 | 2.7 | 0.7 | 100.0 |
Burden of financing with non-defense, non-Social Security cuts | 17.4 | 21.3 | 20.2 | 18.0 | 8.9 | 4.9 | 5.0 | 4.4 | 100.0 |
Source: Author's analysis using data from the Tax Policy Center, Congressional Budget Office and Survey of Consumer Finances, as described in text
The sixth and seventh rows of Table 1 shows the distribution of the TCJA tax cuts by income percentile class, for both the House and Senate versions of the bill. As has been pointed out before, the benefits of both versions of this tax cut are astoundingly regressive, with the overall top 1 percent claiming 47.6 and 62.6 percent of total benefits in the House and Senate versions respectively, while the bottom 40 percent of tax units claim less than 2 percent of the total benefits of the House bill and actually see tax increases from the Senate bill.
But as many analysts have noted, examining only the direct effect of tax cuts without accounting for how these cuts will be financed in the long-run is likely to lead to wildly misleading conclusions about their effects. Given this, Table 1 also shows the distributional impact by income percentile class of two illustrative ways that the tax cut could be financed.
The first financing method examined is simply be a lump-sum tax increase on tax units – every tax unit paying an identical dollar sum large enough to neutralize the economy-wide, overall cost of the TCJA (this lump-sum financing was examined by TPC). It could be objected that this is quite unrealistic: lump-sum taxes don’t exist in the real world and won’t be used to finance the tax cut. This is fair enough – but the entire economic argument that the TCJA will lead to economic growth implicitly presumes that the taxes cut in the framework will be offset by financing that has no behavioral effect at all – which can essentially only occur with lump-sum tax increases. The distribution of the burden of financing the TCJA with a lump-sum increase is predictably regressive. The bottom two-fifths would bear 50.1 percent of the burden while the top 1 percent would face just 0.7 percent.1
Finally, Table 1 shows the distributional impact by income group of financing the TCJA tax cuts with across-the-board cuts in all federal spending except Social Security and defense. This is a rough but fair characterization of the budget resolution passed last month by the Senate and House – very large cuts to all spending except these programs. In fact, the budget resolution, if anything, directed cuts more regressively than an across-the-board cut in non-defense, non-Social Security spending might indicate. For example, cuts to Medicaid fall much more heavily on the bottom of the income distribution than do cuts to Medicare. But Medicare remains a larger overall expenditure, yet the budget resolution directed larger cuts to Medicaid than Medicare.
Yet even with an assumption that actually lightens the likely regressivity of the Congressional budget resolution, this method of financing the TJCA would still result in a steeply regressive burden, with the bottom two-fifths bearing 38.7 percent of the cut (as opposed to their income share of just 13.1 percent) just while the top 1 percent faces just 4.4 percent (relative to their 15.3 percent income share).2
The information in this table allows us to show the total share of income, tax cuts and the burden of financing them that is faced by families by income percentile and race. Table 2 shows the share of total income, the share of tax cuts and the share of the burden of financing these tax cuts facing white, African American and Hispanic families in each income percentile class. The first block of rows simply shows the shares of total income accounted for by race and income percentile. So, for example, white families in the bottom fifth of the income distribution account for 2.6 percent of total income, while African American and Hispanic families in the bottom fifth account for 1.1 and 0.5 percent of total income, respectively. White families in the top 1 percent account for 14.1 percent of total income, while African American and Hispanic families in the top 1 percent account for 0.2 and 0.3 percent of total income (shares that are low because so few African American and Hispanic families are in the overall top 1 percent).
Distribution of overall income, TCJA benefits and burden of financing, by income and racial groups
Overall income | |||||||||
---|---|---|---|---|---|---|---|---|---|
Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total | |
White | 2.6 | 6.0 | 10.6 | 16.5 | 11.9 | 8.4 | 11.5 | 14.1 | 81.5 |
African American | 1.1 | 1.4 | 1.7 | 1.9 | 1.1 | 0.5 | 0.3 | 0.2 | 8.2 |
Hispanic | 0.5 | 1.1 | 1.3 | 1.5 | 0.6 | 0.3 | 0.3 | 0.3 | 5.9 |
Other | 0.2 | 0.3 | 0.5 | 0.8 | 0.7 | 0.6 | 0.8 | 0.6 | 4.4 |
Total | 4.4 | 8.7 | 14.2 | 20.7 | 14.2 | 9.8 | 12.9 | 15.3 | 100.0 |
Benefits of TCJA, House | |||||||||
Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total | |
White | 0.2 | 0.9 | 6.5 | 12.6 | 4.6 | 1.4 | 17.0 | 44.1 | 90.3 |
African American | 0.1 | 0.2 | 1.1 | 1.5 | 0.4 | 0.1 | 0.4 | 0.6 | 2.6 |
Hispanic | 0.0 | 0.2 | 0.8 | 1.1 | 0.2 | 0.0 | 0.4 | 1.0 | 3.0 |
Other | 0.0 | 0.0 | 0.3 | 0.6 | 0.3 | 0.1 | 1.2 | 1.9 | 4.1 |
Total | 0.3 | 1.3 | 8.8 | 15.8 | 5.5 | 1.6 | 19.1 | 47.6 | 100.0 |
Benefits of TCJA, Senate | |||||||||
Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total | |
White | -0.6 | -0.2 | 2.2 | 5.8 | 6.0 | 4.9 | 14.0 | 58.0 | 90.1 |
African American | -0.2 | 0.0 | 0.4 | 0.7 | 0.5 | 0.3 | 0.4 | 0.8 | 2.8 |
Hispanic | -0.1 | 0.0 | 0.3 | 0.5 | 0.3 | 0.2 | 0.3 | 1.3 | 2.7 |
Other | 0.0 | 0.0 | 0.1 | 0.3 | 0.3 | 0.3 | 1.0 | 2.5 | 4.4 |
Total | -1.0 | -0.3 | 2.9 | 7.3 | 7.2 | 5.7 | 15.7 | 62.6 | 100.0 |
Burden of lump-sum financing | |||||||||
Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total | |
White | 16.5 | 15.2 | 14.6 | 13.1 | 6.0 | 2.9 | 2.4 | 0.6 | 71.4 |
African American | 6.7 | 3.5 | 2.4 | 1.5 | 0.5 | 0.2 | 0.1 | 0.0 | 15.0 |
Hispanic | 3.5 | 2.8 | 1.8 | 1.2 | 0.3 | 0.1 | 0.1 | 0.0 | 9.7 |
Other | 1.2 | 0.7 | 0.7 | 0.7 | 0.3 | 0.2 | 0.2 | 0.0 | 3.9 |
Total | 27.9 | 22.1 | 19.6 | 16.5 | 7.1 | 3.4 | 2.7 | 0.7 | 100.0 |
Burden of financing by cuts in non-defense, non-Social Security spending | |||||||||
Bottom Fifth | Second Fifth | Middle Fifth | Fourth Fifth | 80th-90th percentile | 90th-95th percentile | 95th-99th percentile | Top 1 percent | Total | |
White | 10.3 | 14.6 | 15.1 | 14.3 | 7.4 | 4.2 | 4.5 | 4.0 | 74.5 |
African American | 4.2 | 3.4 | 2.5 | 1.7 | 0.7 | 0.3 | 0.1 | 0.1 | 12.8 |
Hispanic | 2.2 | 2.7 | 1.9 | 1.3 | 0.4 | 0.1 | 0.1 | 0.1 | 8.7 |
Other | 0.7 | 0.7 | 0.7 | 0.7 | 0.4 | 0.3 | 0.3 | 0.2 | 4.0 |
Total | 17.4 | 21.3 | 20.2 | 18.0 | 8.9 | 4.9 | 5.0 | 4.4 | 100.0 |
Source: Author's analysis using data from the Tax Policy Center, Congressional Budget Office, and Survey of Consumer Finances, as described in text.
The next two blocks show the shares of the TCJA tax cuts by race and income percentile, for both the House and Senate versions. For the House bill, the first row in this block of columns, for example, shows that 0.2 percent of the benefits of the House TCJA go to white families in the bottom fifth of the income distribution, 0.1 percent of the benefits go to African Americans in the bottom fifth, and less than 0.005 percent goes to Hispanic families in the bottom fifth.
Moving to the top 1 percent row, this shows that 44.1 percent of the benefits of the House TCJA go to white families in the top 1 percent, 0.6 percent go to African American families in the top 1 percent, and 1.0 percent go to Hispanic families in the top 1 percent. The final row sums these to show total benefits of the tax cut by race, but it is important to realize that (for example), while 90.3 percent of the total benefits go to white families, this is driven overwhelmingly by just the top 1 percent.
For the Senate bill, the bottom 40 percent of all races see tax increases from the TCJA. Moving to the top 1 percent row, 58.0 percent of the benefits of the Senate TCJA go to white families in the top 1 percent, 0.8 percent go to African American families in the top 1 percent, and 1.3 percent go to Hispanic families in the top 1 percent. The final row sums these to show total benefits of the tax cut by race, but it is important to realize that (for example), while 90.1 percent of the total benefits go to white families, this is driven overwhelmingly by just the top 1 percent.
The next group of rows shows the distribution of the burden of financing these tax cuts if done with lump-sum taxes, by both income percentile class and race. Again, the first row can be read to show that white families in the bottom fifth would bear 16.5 percent of the total burden of financing tax cuts with lump-sum increases, while African American and Hispanic families in the bottom fifth would bear 6.7 and 3.5 percent, respectively. White families in the top 1 percent would bear just 0.6 percent of the total cost of financing tax cuts with lump-sum tax increases, while African American and Hispanic families would bear less than 0.05 percent of the burden. These numbers on lump-sum tax burden do not perfectly reflect population shares because there are more tax units in the lower fifths of the population than in higher fifths. This, combined with higher African American and Hispanic shares in these lower fifths, means that a lump-sum tax levied on tax units will hit them disproportionately.
Finally, the fourth set of bars shows the distribution of the burden of financing these tax cuts if it was done with an across-the-board cutback in government spending, except for defense and Social Security. Again, despite making an assumption that softens how regressive the spending cut in the Congressional budget resolution would be, the burden still falls heavily on non-white families because of the overall regressivity. White families in the bottom fifth would bear 10.3 percent of the burden, while African American and Hispanic families in the bottom fifth would bear 4.2 and 2.2 percent. White families in the top 1 percent would bear just 4 percent of the burden, while African American and Hispanic families in the top 1 percent would both bear 0.1 percent of the burden.
The regressive distribution of both benefits from the TCJA and the burden of financing them is near-entirely a function of income ranking. But we know that the historical legacy of racism and discrimination has led to a non-random distribution of racial groups across the income distribution, with the result that white families overall would see larger benefits from tax cuts and would bear lighter burdens from financing them, while non-white families see smaller gains but larger burdens from the TCJA. To be clear, the benefits even across white families are extraordinarily concentrated at the top – a third of white families will see outright tax increases from the Senate bill, for example. But we shouldn’t be blind to disproportionate racial costs and benefits from fiscal policies.
1. Calculating the burden of financing tax cuts with lump-sum tax increases is done by simply taking Tax Policy Center (TPC) estimates of the share of total tax units in each income percentile class. The lower fifths of the income distribution tend to have more tax units (i.e. relatively more single filers), so, a lump-sum tax levied on tax units would fall more heavily on lower-income tax units.
2. Calculating the burden of financing tax cuts with across-the-board cuts in government spending except Social Security and Defense is done by using data from the Congressional Budget Office (CBO). The CBO provides data that allows estimates on the distribution of total transfer payments except for Social Security by income percentile class. CBO also provides guidance on how to distributionally allocate the benefit of federal government consumption and investment spending (or non-transfer spending). The CBO suggest two methods – allocating the benefit of this as a lump-sum across the population, or allocating benefits proportional to income. We allocate non-transfer spending by taking a simple average of these two approaches. We then weight the transfer and non-transfer cuts by current shares of transfers and non-transfer spending in federal spending. In 2016, transfers to persons besides Social Security were roughly half of all non-defense federal spending.
Enjoyed this post?
Sign up for EPI's newsletter so you never miss our research and insights on ways to make the economy work better for everyone.