chevron-down Created with Sketch Beta.
August 02, 2018 Practice Point

How the Byrd Rule Might Have Killed the 2017 Tax Bill . . . and Why It Didn’t

By George K. Yin, Edwin S. Cohen Distinguished Professor of Law and Taxation, University of Virginia, Charlottesville, VA

During debate on the 2017 tax bill that was once slated to become the “Tax Cuts and Jobs Act,” the number of stories referencing an arcane budget law known as the “Byrd rule” skyrocketed in publications as diverse as the New York Times and the weekly trade magazine, Tax Notes.1 Many of the references, however, dealt with minor effects of the rule, such as its role in removing the bill’s short title from the final legislation. This article briefly describes a much more important and little-known aspect of this little-known rule that might have—and perhaps should have—killed the bill altogether.

Preventing Out-Year Deficit Increases

The Byrd rule potentially prevents Senate consideration of any part of a reconciliation bill or conference report that is considered to be “extraneous” to the underlying bill. The rule creates a point of order that may be raised by any Senator against any provision in such legislation. If the point of order is sustained and not waived by at least three-fifths of the Senate, the offending provision must be removed from the bill.2

An “extraneous” provision includes one that would increase the deficit in any fiscal year—sometimes termed an “out-year”—after the budget period covered by the reconciliation bill. This part of the rule arose in 1987 when reconciliation bills were still used only for deficit reduction purposes. Some Senators tried to elude the deficit reduction objective by including provisions that raised revenue during the period covered by the bill (thus helping to meet the bill’s deficit reduction target) but lost revenue in an out-year. The Senate curbed this practice by generally defining as extraneous any provision increasing the deficit in an out-year.3 Many attribute this portion of the Byrd rule as the reason why virtually all major tax cuts enacted through reconciliation since 2001 have been sunset.

Yet the single biggest tax cut included in the 2017 tax legislation—the dramatic lowering of the corporate tax rate—was not sunset. That provision was estimated to lose about $1.35 trillion during the legislation’s ten-year budget period (2018-2027)4 and, because it was enacted as a “permanent” provision, will continue to increase the deficit after 2027. So how did it escape a Byrd rule objection? Before answering, let us consider one other important piece of background information.

Objections to Provisions in Conference Reports

A Byrd rule point of order can be raised during the Senate’s consideration of a conference report of a reconciliation bill. But a conference report—the product of a conference committee’s compromise of differences in House and Senate bills on a piece of legislation—is a special class of legislation. To achieve finality in the legislative process, a conference report is generally subject to only an up-or-down vote in the House and Senate. No more amendments are allowed, or else the process would be endless.

So what happens if a Byrd objection is sustained against a provision in a conference report being considered by the Senate? In that instance, the law permits the conference bill to be amended by removing the offending provision but making no other change.5 The conference bill without the offending provision is then considered by the Senate and, if approved, sent back to the House for its up-or-down vote.

Now put those two pieces of information together. Suppose an objection had been raised and sustained against the corporate tax rate cut included in the 2017 conference report. At that point, the only option would have been for the Senate and House to vote up or down on the legislation without the corporate rate cut.

What would have happened? I believe support for the entire bill would have collapsed. Without the corporate rate cut, few members of Congress would have supported the corporate tax base broadeners remaining in the bill, such as repeal of former section 199. Support for the international tax changes would also probably have disappeared. Finally, there would have been little reason to make most of the other changes in the bill, including enactment of new section 199A and the tax cuts for individuals. This possible scenario shows the potential significance of the Byrd rule. A single objection permitted by the rule, if sustained and supported by at least 41 Senators, could have scuttled—without use of the filibuster—a major policy initiative of a majority in Congress.

The timing of the objection would have been crucial. Many stories in the media and among Hill observers dealt with the possible role of the Byrd rule during the Senate’s consideration of its own bill. But the same objection raised and sustained at that stage of the legislative process would not have had the same effect because the Senate could have amended its bill. The obvious change would have been to sunset the corporate rate cut after ten years. Supporters would not have liked that change, but it surely would have been better than no corporate rate cut at all. Only if the Byrd objection had been raised at the conference stage—when it would have been too late to make any other amendment—could it have had its dramatic effect of killing the bill.

So—why wasn’t an objection raised against the corporate tax rate cut in the conference report? And if it was raised, why wasn’t it sustained?

Out-Year Deficit Increases Determined on a “Net” Basis

Part of the answer is that a Byrd rule violation is determined on a “net” basis. If the amount a provision increases the deficit in an out-year is offset by enough deficit reduction in that year from other provisions, then there is no violation.

When the Senate considered its bill, supporters worked to ensure that it would be Byrd-compliant. The major change made to the House bill was to sunset the tax cuts for individuals after 2025 to eliminate their revenue loss in the out-years. But supporters made other changes as well, including a curious one to delay a House provision requiring amortization over five years of certain previously deductible R&E expenditures. The House had proposed making it effective beginning in 2023 but the Senate proposed to delay it until 2026, with the following budgetary consequence to the Senate bill:

Selected Revenue Effects of 2017 Tax Legislation (Senate Bill)
(billions of dollars)

corporate tax rate cut
2018-2024: . . .
2025: -157
2026: -163
2027: -171
after 2027: greater losses

amortization of R&E expenditures
2018-2024: . . .
2025: -0-
2026: +26
2027: +36
after 2027: smaller gains

Net total
2018-2024: . . .
2025: -156
2026: -59
2027: +34
after 2027: ???

Source: Jt. Comm. on Tax’n, Estimated Revenue Effects of the “Tax Cuts and Jobs Act,” as passed by the Senate on Dec. 2, 2017, JCX-63-17, items II.A.1, II.C.9, and net total.

The Senate may have delayed the R&E provision at the insistence of the affected taxpayers. But, as the table shows, delay also had an important budgetary effect by helping to make the overall Senate bill a revenue raiser in 2027. Without the estimated $36 billion produced by the R&E provision in that year, the overall bill would have resulted in a small revenue loss in 2027.

But what did the bill’s budget effect in 2027—the last year of the bill’s budget period—have to do with the Byrd rule, which is concerned with budget effects in years after that period (i.e., after 2027)? One explanation is that it supported a possible claim that since the bill reduced the deficit in 2027, it would do likewise in every year after 2027 (and hence not violate the Byrd rule).

At least one news report bought this claim.6 Tax professionals, however, know that it is not necessarily true. The budget effect of some provisions changes and even reverses from one year to the next.

Indeed, the R&E change was one such provision. The revenue estimate of the provision in the House bill showed that its second year would be its peak year for raising revenue.7 After that year, more of the revenue gain from amortizing (and not deducting) new expenditures would be offset by revenue loss from the continued amortization of prior expenditures, thereby causing the provision to produce less than $36 billion per year of net revenue gain after its second year (i.e., after 2027 in the Senate bill). The last column in the table identifies this effect, leaving uncertainty as to the budget impact of the entire bill in the out-years.

Congress changed these and other provisions in the conference report, but this example reveals an important omission in the analysis of the bill. Since the corporate rate cut will increase the deficit in the out-years, the only way it can avoid being a Byrd violation is if there are sufficient offsets, but do we know the budget effect of the entire bill after 2027? Remember that the JCT estimated that one of the bill’s biggest revenue raisers—the taxation of previously deferred foreign income—would not raise any revenue after 2026.8 Remember also that the Byrd rule is violated if in any single out-year, the overall bill increases the deficit. If so, a provision like the corporate rate cut contributing to that deficit increase could have been struck from the bill.

But aren’t these simply questions to be resolved by budget economists? After all, there were other provisions in the final bill, including repeal of the Affordable Care Act’s individual mandate penalty and especially the change in inflation indexing, that could be expected to reduce the deficit in years after 2027. Maybe their budget effect would more than offset the out-year deficit increases produced by the rest of the bill. There are, however, at least two problems with that answer.

Determining Byrd Rule Compliance   

One problem is practical. Estimated budget effects of proposed legislation are measured against a baseline of what would have transpired had there been no change in law. The JCT relies upon baseline parameters provided by the CBO, but the CBO generally doesn’t provide such parameters beyond ten years. In general, if there is no baseline, there can be no estimates, and if there are no estimates, there is no way to determine if a Byrd rule violation has been avoided.

A panelist at a recent tax conference stated that, notwithstanding the absence of this information, the JCT had estimated that the inflation indexing change would raise a large amount of revenue during the first ten out-years (2028-2037). This estimate—not found by me on any of the pertinent congressional websites—would not be dispositive of the Byrd rule question, which requires an estimate of the budget consequences of the entire bill for a potentially unlimited number of future years. One out-year estimate that is publicly available is a statement from the CBO and JCT—in response to a different congressional budget law requirement—that the bill “would not increase on-budget deficits by more than $5 billion in any of the four consecutive 10-year periods beginning in 2028.”9 This statement is obviously also not dispositive of the Byrd question.

But there is a more fundamental problem with determining compliance with the Byrd rule. As illustrated by the foregoing CBO and JCT statement, skilled economists may be able to make reasonable assumptions to support estimates of long-term budget consequences. But the Byrd rule doesn’t require reliance on such economists. Rather, the law specifies that all budget determinations are to be made by the House and Senate budget committees, meaning, effectively, the chairs of those committees.10 Those chairs may—but apparently need not—seek out the professional expertise I have described.

We know this because of one final part of the story. In 2006, the Senate was considering a reconciliation bill that extended the reduced tax rates for dividends and long-term capital gain. The bill covered only five years, and the estimates for the years after the budget period—in this case, years 6-10—showed that it lost revenue. In other words, the bill was clearly in violation of the Byrd rule.

Congress overcame this problem by adding to the bill a provision allowing more taxpayers to convert their regular IRAs to Roth IRAs. The JCT estimated that conversion would produce enough additional revenue in years 6-10 to offset the revenue loss from the bill in those years.11 But the JCT also estimated that over ten years, the conversion provision would lose revenue.12 Despite this, and despite protests from members of the minority party in Congress, the Senate budget committee chair refused to certify that the bill as amended still violated the Byrd rule.13 In other words, the addition of a new tax cut to a bill already in violation of the Byrd rule because of its revenue loss miraculously eliminated the problem. The budget committee chair, who had carefully cultivated a reputation of fiscal responsibility over the years, flinched when he finally had an opportunity to act on his purported beliefs.


So there you have it. A determination of a Byrd rule violation requires analysis by those with budget estimating expertise, yet the law does not mandate that consultation. The rule is therefore an example of a potentially powerful tool that can be incapacitated by the way it is implemented. It is an all too familiar phenomenon in Congress. If the reconciliation process is now to be used principally to enable thin majorities in Congress to pass important legislative priorities—without regard to the budget effect of the legislation—it is not likely that a senior member of such a majority, such as a budget committee chair, will thwart that goal by requiring strict compliance with the Byrd rule.

Should Congress ever decide to take deficit reduction seriously—and is it not finally about time?—it would be a simple fix to make the Byrd rule administrable and consequential by requiring reliance on specific, published analysis provided by the CBO or JCT. Until then, the Byrd rule may be just another legislative frill—often a nuisance, and maybe sometimes a blessing, but never anything that would prevent a congressional majority from getting its way.

1 See Ellen P. Aprill & Daniel J. Hemel, The Tax Legislative Process: A Byrd’s Eye View, 81 Law & Contemp. Probs. 99, 99-100 (2018).

2 See Congressional Budget Act of 1974, as amended, Pub. L. No. 93-344, § 313, 88 Stat. 297, 2 U.S.C. § 644 (2012).

3 See 2 U.S.C. § 644(b)(1)(E); George K. Yin, Temporary-Effect Legislation, Political Accountability, and Fiscal Restraint, 84 N.Y.U. L. Rev. 174, 218 (2009).

4 See Jt. Comm. on Tax’n, Estimated Budget Effects of the Conference Agreement for H.R. 1, the “Tax Cuts and Jobs Act,” JCX-67-17, item II.B.

5 See 2 U.S.C. § 644(d).

6 See Jim Puzzanghera & Lisa Mascaro, Changes May Cost Tax Bill GOP Votes, L.A. Times (Nov. 16, 2017) at A13 (stating that revenue gain in 2027 “allow[ed] the bill to conform with the Byrd rule”).

7 See Jt. Comm. on Tax’n, Estimated Revenue Effects of H.R. 1, the “Tax Cuts and Jobs Act,” as ordered reported by the Committee on Ways and Means on Nov. 9, 2017, JCX-54-17, item II.E.15.

8 See JCX-67-17, supra n. 5, item III.A.3.

10 See 2 U.S.C. § 643(a); cf. 2 U.S.C. § 641(d)(4).

11 See Jt. Comm. on Tax’n, Estimated Revenue Effects of the Conference Agreement for the “Tax Increase Prevention and Reconciliation Act of 2005,” JCX-18-06, item V.10 and net total.

12 See 152 Cong. Rec. 8014-8015 (May 11, 2006) (reproducing JCT 10-year revenue estimate).

13 See id. at 8018-8019 (statement of Sen. Gregg (R.-N.H.)). For more detail on this episode, see Yin, supra n. 4, at 221-24.