The statute whose name is not its name

Every American who followed the news in December 2017 knows the name of the largest tax rewrite in a generation. The Tax Cuts and Jobs Act cut the corporate rate, nearly doubled the standard deduction, capped the deduction for state and local taxes, and rewrote the rules for taxing income earned abroad. There is only one problem with that familiar name. It is not the name of the law. Days before the bill reached the president’s desk, a Senate point of order struck the short title from the text, and the measure was enrolled under a long descriptive title that almost nobody has ever quoted. The statute everyone calls the Tax Cuts and Jobs Act is formally “An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018.” The popular name survived in headlines and conversation. It did not survive in the law.

That single episode tells the reader most of what this profile exists to explain. The 2017 tax statute was not written the way most landmark laws are written. It moved through Congress on the budget reconciliation track, which let it clear the Senate with a simple majority but subjected every line of it to a procedural screen called the Byrd rule. The Byrd rule is the reason the individual tax cuts carry expiration dates while the corporate rate cut does not. It is the reason the bill’s own name was deleted before enactment. And it is the reason the statute looks the way it does: a collection of measures whose durability was decided not by any judgment about which policies deserved to last, but by whether each item’s cost fit inside a ten-year budget instruction. That is the reconciliation asymmetry, and it is the key to the entire law.

This article is a complete profile of Public Law 115-97, signed on December 22, 2017. It carries the passage, the provisions, the procedure and the evidence in one place, because this statute has no specialist siblings in the series and the profile must do all four jobs. A reader who works through it will be able to explain why the individual cuts expire while the corporate cut does not, state the procedural rule that forced that asymmetry, and recite the fact that almost no coverage reports: the law is not formally named the Tax Cuts and Jobs Act at all. Those three abilities are the article’s test, and everything that follows is organized to make them attainable. The profile proceeds in order: the road to the bill, the passage and the conference that produced it, the Byrd bath that struck the title, the procedural asymmetry that explains the statute’s shape, the four provision groups, the implementation record, the permanent-or-expiring table that collects them, the revenue and distributional evidence, the investment and wage literature with the growth debate, the comparison with 1986, and the verdict.

The scale of the statute justifies the length of the treatment. The Joint Committee on Taxation, the official estimator for tax legislation, scored the measure as a net federal revenue loss of $1.456 trillion over fiscal years 2018 through 2027, the largest tax reduction enacted since the early 1980s. It reset the brackets and the standard deduction for individuals, capped two of the most widely used itemized deductions, doubled the estate tax exemption, and zeroed out the penalty for going without health coverage. It cut the corporate rate from 35 percent to 21 percent permanently, repealed the corporate alternative minimum tax, let firms write off new investment immediately for a time, limited the deduction for business interest, and created a new 20 percent deduction for pass-through income. And it rebuilt the taxation of cross-border income around a territorial principle, with a one-time tax on old foreign earnings, new minimum taxes on intangible income earned abroad, and a backstop against base erosion.

The procedure deserves equal billing with the substance, because the procedure is the substance in a sense that applies to few other statutes. Most landmark laws are compromises between policy goals. This one was also a compromise between policy goals and a Senate rulebook. Reconciliation is a fast-track process created by the Congressional Budget Act that lets budget-related legislation pass the Senate with 51 votes instead of 60, at the price of strict limits on what the legislation may contain. The price is collected by the Byrd rule, named for Senator Robert Byrd of West Virginia, which lets any senator strike provisions deemed extraneous to the budget task. One of its tests bars any provision that increases the deficit beyond the ten-year budget window. That single test is the reason the drafters gave the individual provisions expiration dates: the sunsets kept the official score inside the $1.5 trillion limit set by the fiscal year 2018 budget resolution, and without them the bill would have failed the rule and died on the floor. The corporate rate needed no sunset because its cost fit the window as scored. Procedure determined text, and in this statute the determination is visible in almost every section. That is the sharpest and most recent appearance in this series of the thesis that procedure writes law, and the 2017 act states it more plainly than any earlier statute the series has covered.

A word on naming before the profile begins. Throughout this article, “the 2017 act” and “the statute” refer to Public Law 115-97 as enacted. “The Tax Cuts and Jobs Act” appears where the popular name is the subject, because readers search for that name and deserve to find this profile. But the article never presents the popular name as the enacted short title, and neither should any citation. The enrolled law carries the descriptive reconciliation title quoted above, and the difference between the name everyone uses and the name the law bears is one of the four near-misses this profile is built to prevent. The other three are equally common: the assumption that the whole act is permanent, the assumption that the sunsets reflect a policy judgment rather than a procedural requirement, and the confusion of the conference rate of 21 percent with the 20 percent figure in the pre-conference bills. Each will be marked where it most often misleads.

Why 2017: the road to the bill

No statute arrives without a past, and the 2017 act’s past explains why its drafters reached for the tools they did. The last comprehensive rewrite of the federal tax code had been the Tax Reform Act of 1986, thirty-one years earlier. In the decades between, Congress had layered provision upon provision onto the code: temporary extenders renewed year after year, targeted credits and deductions added for favored activities, and rate schedules adjusted piecemeal. By 2017 the code was longer, more complex and more visibly distorted than at any point since 1986, and the case for another comprehensive rewrite had been made, in some form, by nearly every president and every tax-writing chairman of the intervening generation. What had been missing was a vehicle and a moment.

The corporate tax supplied the urgency. The federal statutory rate of 35 percent, combined with state corporate taxes, gave the United States the highest statutory corporate rate in the developed world, a fact cited in nearly every reform proposal of the period. High statutory rates do not by themselves prove that firms pay high effective rates, and American multinationals proved adept at lowering theirs, but the headline rate shaped behavior at the margin. It encouraged a string of corporate inversions, transactions in which American firms reincorporated abroad to escape the U.S. tax net, and it prompted successive Treasury rulemakings aimed at blocking them, including rules that helped sink one of the largest proposed mergers of the decade. It also sustained the lockout effect: because foreign earnings were taxed only when brought home, American multinationals accumulated trillions of dollars in untaxed earnings in foreign subsidiaries and kept them there, a stockpile that grew with every profitable year. The worldwide system with deferral had become, in practice, a system that taxed foreign earnings at a rate of zero for any firm patient enough never to repatriate. Reformers of both parties agreed this was broken. They disagreed about what should replace it.

Congress had rehearsed the rewrite more than once. In 2014 the House Ways and Means chairman released a comprehensive reform draft proposing lower rates, a territorial system and a bank tax; it was praised as a technical achievement and then shelved. In 2016 the House majority published a tax blueprint floating a destination-based cash-flow tax with border adjustment, which proved too controversial to survive contact with retailers, importers and the Senate, and it was abandoned before the 2017 bill was drafted. The border-adjustment fight showed the limits of what the reconciliation vehicle could carry: a fundamental redesign of the tax base, with uncertain revenue consequences and certain political enemies, that the Senate majority concluded could not survive the Byrd bath or the score. The international architecture of the enacted statute, the participation exemption paired with anti-abuse backstops, was the descendant of ideas worked out in the failed drafts, stripped of the border adjustment and fitted to the reconciliation score.

The moment arrived with the 2016 elections, which gave one party unified control of the House, the Senate and the presidency for the first time in a decade. Tax reform had been a stated priority of the new majority’s leadership for years, and after an earlier health-care legislative effort did not advance, the tax bill became the vehicle into which the majority poured its ambitions. The choice of H.R. 1 as the bill number, the designation traditionally reserved for the majority’s top priority, signaled the intent. The bill was introduced on November 2, 2017, marked up within days, and passed by the House two weeks later. The speed was not an accident or a shortcut. It was the reconciliation timetable operating as designed, and it meant the statute would be drafted, scored, amended and passed in roughly seven weeks. Everything distinctive about the law, its sunsets, its struck title, its 21 percent rate, was produced under that compression.

One more piece of context belongs here, because it explains a provision that otherwise looks out of place. The individual mandate’s penalty was zeroed in a tax bill, which struck many observers as odd: what did health coverage have to do with tax reform? The answer is reconciliation arithmetic. Zeroing the penalty was scored as reducing federal outlays on health coverage subsidies, and those savings helped the bill fit inside the $1.5 trillion instruction. The provision entered the bill during Senate consideration, where the budget math was tightest, and it survived the Byrd bath because it plainly affected revenues and outlays. Policy coherence did not put the mandate provision in the tax bill. The score did. It is a small example of the profile’s large claim, and the reader who grasps it has grasped the statute’s logic.

Fifty days from introduction to signature

The bill that became Public Law 115-97 was introduced in the House as H.R. 1 on November 2, 2017, by Representative Kevin Brady of Texas, then chairman of the Ways and Means Committee. The committee marked it up and reported it within a week, the full House passed it on November 16 by 227 to 205, and fifty days after introduction the president signed it. Between those bookends the measure moved with a speed that regular-order legislation almost never achieves, because it was never on the regular-order track. It was a reconciliation bill, and reconciliation sets its own tempo.

The committee stage, compressed as it was, still left its mark on the bill. The Ways and Means markup ran four days and reported the bill on November 9 by 24 to 16, largely along party lines, with amendments that refined the pass-through provisions and the international title. In the Senate, the Finance Committee under Chairman Orrin Hatch of Utah marked up its own version, which differed from the House bill in the respects the conference later reconciled: the Senate version kept the individual alternative minimum tax, set a different mortgage ceiling, and, notably, added the mandate-penalty zeroing that the House bill had not contained. The Senate version also retained the full $1,000,000 mortgage ceiling that the conference would later cut to $750,000. Each chamber’s bill was thus a distinct reconciliation product, shaped by its own committee’s reading of the score, and the conference committee’s job was to merge two scores into one. The conference report filed December 15 did exactly that, and the Byrd bath then scrubbed the merged product a final time before passage.

The vehicle matters more than the calendar. In October 2017 the House adopted the Senate-passed fiscal year 2018 budget resolution, H. Con. Res. 71, by 216 to 212, and that resolution carried reconciliation instructions directing the tax-writing committees to produce legislation reducing revenues by not more than $1.5 trillion over fiscal years 2018 through 2027. The resolution had passed the Senate on October 19 by 51 to 49, with the same narrow margin that would later pass the bill itself, and the House’s concurrence a week later completed the procedural foundation. The $1.5 trillion figure was itself a negotiated compromise: deficit-conscious senators wanted a smaller number, the tax-writing committees wanted room to draft, and $1.5 trillion was the number that could command 51 votes. Everything downstream, the rate, the sunsets, the struck title, flowed from that number. Those instructions were the bill’s license to move. Under reconciliation, Senate debate is limited, amendments are restricted, and the filibuster is unavailable, which means a determined majority can pass budget-related legislation without a single vote from the minority. The trade is that the Byrd rule polices the contents, and any senator can raise a point of order against provisions that are extraneous to the budget purpose, including anything that raises the deficit outside the ten-year window. The 2017 bill was drafted inside that cage from its first page.

How did the bill pass the Senate with only 51 votes?

Reconciliation let the bill move under the fiscal year 2018 budget resolution, which barred a filibuster and set the passage threshold at a simple majority. With 51 votes the majority could pass the measure without any minority-party support, and every Senate passage vote cleared that bar without a vice-presidential tie-break.

The Senate’s first passage vote came in the early hours of December 2, 2017, after a marathon amendment session known as a vote-a-rama. The tally was 51 to 49, with Senator Bob Corker of Tennessee the only Republican voting no, on deficit grounds. No vice-presidential tie-break was needed, because 51 votes is itself a majority of the Senate. The session produced the only vice-presidential tie-break of the entire process, and one persistent error deserves correction here: Vice President Pence did cast a tie-breaking vote, but it was on an amendment by Senator Ted Cruz concerning tax-preferred education savings accounts, decided 51 to 50, and not on passage of the bill. The Cruz amendment was one of the three provisions the Byrd bath would strike two and a half weeks later, which means the session’s only tie-break was cast for a provision that never became law.

The conference committee reconciled the House and Senate versions and filed its report, H. Rept. 115-466, on December 15. The House adopted the conference report on December 19 by 227 to 203. That vote, however, was taken on a version of the bill that still contained three provisions the Senate parliamentarian had advised were impermissible under the Byrd rule. On the Senate floor that same day, Senator Bernie Sanders of Vermont, the ranking member of the Budget Committee, raised the points of order against all three, in a joint effort with Senator Ron Wyden of Oregon, the ranking member of the Finance Committee. Senator Mike Enzi of Wyoming moved to waive the rule, which requires 60 votes, and the waiver failed 51 to 48. The three provisions were stricken, and the Senate passed the amended bill late on December 20 by 51 to 48, with Senator John McCain of Arizona absent, in Record Vote 323. Because the Senate had changed the bill, the House had to vote again, and it concurred in the Senate amendment on December 20 by 224 to 201, in Roll 699. The president signed the enrolled bill on December 22, 2017, and it became Public Law 115-97. The title-striking sequence is told in full in the Byrd bath section below.

Two features of that sequence deserve emphasis. First, the title-striking is sometimes attributed to the parliamentarian alone. The parliamentarian advises; senators act: it was Senator Sanders who raised the points of order, on the parliamentarian’s advice, and the failed waiver vote removed the provisions. Second, the final tallies are the ones that belong in any citation: 51 to 48 in the Senate and 224 to 201 in the House, both on December 20. The 227 to 203 House vote of December 19 belongs to the version that never became law.

The reconciliation vehicle is worth understanding in its own right, because the 2017 act is the purest example in the series of what the vehicle does to legislation. For the full mechanics of the process, see the reconciliation process explained. What matters here is the bargain: speed and a simple-majority threshold in exchange for the Byrd rule’s policing of content. The 2017 drafters accepted the bargain and then designed the entire statute around the policing. Every sunset date in the law is a receipt for that bargain, and the struck title is its most visible stamp.

The House and Senate bills that went into conference differed in ways that reveal what each chamber valued and what the score would tolerate. The House bill set the mortgage interest ceiling at $500,000 for new debt, a figure that survives in some secondary accounts and must not be mistaken for law; the Senate bill kept the $1,000,000 ceiling; the conference split the difference at $750,000. The House bill repealed the individual alternative minimum tax outright; the Senate retained it; the conference retained it with the higher exemptions. Both chambers’ bills set the corporate rate at 20 percent, and the conference raised it to 21 percent for the scoring room the final agreement needed. The conference report filed December 15 was not a compromise between two philosophies of taxation. It was a compromise between two scores.

The Conference: fourteen days to reconcile two bills

The Senate’s December 2 passage set a fourteen-day clock. The House and Senate versions of H.R. 1 differed on enough provisions to require a conference committee, the joint House-Senate body that reconciles competing texts when neither chamber will simply accept the other’s. Conferees were drawn from the tax-writing committees, Ways and Means for the House and Finance for the Senate, with the majority party controlling the outcome in both delegations. The conference’s formal product was House Report 115-466, filed December 15, and its informal product was the final set of trade-offs that fit the bill inside the $1.5 trillion envelope.

The conference’s most visible decision was the corporate rate. Both chambers’ pre-conference bills had set it at 20 percent. The conference agreement set it at 21 percent, effective for tax years beginning after December 31, 2017, the same effective date the Senate bill had ultimately carried after its one-year delay was dropped. The single point mattered twice over. Substantively, it was the difference between the headline number the bill’s supporters had campaigned on and the number the budget arithmetic allowed. Procedurally, it demonstrated how the envelope disciplined even the bill’s centerpiece: when the conference needed room for the final shape of the individual provisions and the international title, the corporate rate absorbed the adjustment, because a point on the corporate base was the largest single lever available. Anyone repeating the 20 percent figure is describing the drafts, not the law, and the persistence of the error shows how early coverage calcified around numbers the conference later changed.

The conference also settled the individual bracket structure, reconciling the House and Senate approaches into the seven brackets of 10, 12, 22, 24, 32, 35, and 37 percent, and produced the final forms of the international provisions, including the GILTI and FDII deduction percentages and the BEAT rate schedule. On the pass-through question, the conference kept the Senate’s section 199A deduction approach rather than the House’s separate rate structure, a choice that shaped the compliance burden practitioners would face. The conference report was a take-it-or-leave-it package: under the rules, neither chamber could amend it, only vote it up or down. That constraint concentrated enormous bargaining power in the conferees and the leadership, and it meant that the December 19 House vote, 227 to 203, was a vote on the entire compromise at once, including the three provisions the parliamentarian had already flagged as extraneous. The House voted first, before the Senate’s Byrd bath consequences were fully resolved, which is why the title-strike sequence played out the way it did: the House approved a bill with the short title intact, the Senate removed it, and the House had to vote again on the amended text.

The conference’s work also illustrated the asymmetry between the two chambers’ bargaining power. The House wanted its priorities preserved; the Senate, needing to hold fifty of its fifty-two Republicans under reconciliation’s constraints, needed provisions that could survive the Byrd rule. Where the two goals conflicted, the Senate’s constraints won, because the Senate’s floor was where the bill could die. The bill that emerged was the one the budget rules permitted, not the one either chamber would have written alone.

Counting to fifty-one: how a two-vote margin shaped the bill

The bill’s substance cannot be separated from the Senate’s arithmetic, because the arithmetic was the substance’s binding constraint. The Republican majority held fifty-two seats, and reconciliation’s simple-majority threshold meant the bill could lose no more than one Republican vote if every Democrat voted no, with the vice president available as a theoretical fifty-first vote in a 50-50 tie that never materialized on passage. In practice, the bill’s managers operated with a margin of one. Every provision had to satisfy two tests simultaneously: it had to survive the parliamentarian’s Byrd bath, and it had to hold the votes of fifty senators whose individual demands could each be decisive. A single defection beyond Corker would have killed the bill, and every senator knew it.

The December 2 vote showed the margin working. Senator Bob Corker of Tennessee voted no, the only Republican to do so, on the stated ground that the bill added to the deficit, and his defection was absorbed because no other Republican joined him. The bill passed 51 to 49 in the early morning hours. With no vote to spare, the leadership negotiated individual provisions against individual senators’ demands, and the bill’s final shape reflected the overlapping constraints of fifty separate veto points rather than any single coherent design.

The final votes tightened the margin further. On December 20, Senator John McCain of Arizona was absent, and the Senate passed the amended bill 51 to 48, Record Vote 323. The waiver vote on the Byrd rule points of order, which needed sixty votes and failed 51 to 48, demonstrated the other side of the arithmetic: fifty-one votes were enough to pass the bill but not enough to protect its extraneous provisions, and the majority’s inability to reach sixty is what cost the bill its title. The same fifty-two seats produced both outcomes, passage and amputation, because reconciliation sets different thresholds for different questions.

The House’s arithmetic was more comfortable but not without its own drama. The November 16 vote, 227 to 205, included thirteen Republican no votes, concentrated among members from high-tax states who objected to the SALT cap. The December 20 revote, 224 to 201, showed three votes of erosion from the December 19 tally. None of these House votes was ever in doubt the way the Senate’s were, but each recorded the political cost of the bill’s choices, and the party-line character of every roll call meant the bill belonged entirely to its supporters.

The narrow-margin dynamic is the political complement to the reconciliation asymmetry. The procedural rules determined which provisions could last; the vote arithmetic determined which provisions could pass at all. Together they explain why the bill looks the way it does: a maximal tax cut fitted to a $1.5 trillion envelope, passed with one vote to spare, edited by the parliamentarian, and stripped of its name on a point of order.

The Byrd bath and the night the title fell

The mechanism that struck the title is worth understanding in its own right, because it is the same mechanism that wrote the sunset dates. Before a reconciliation bill reaches the Senate floor, its provisions go through what staff call the Byrd bath: the Senate parliamentarian reviews each provision against the six tests of the Byrd rule and advises which ones are extraneous. The parliamentarian’s advice is advisory, but in practice it is decisive, because any senator can raise a point of order against an advisedly extraneous provision, and overcoming the point of order requires sixty votes to waive the rule. In a chamber where the majority held fifty-two seats, sixty votes meant Democratic support, which the bill’s writers knew they would not get. The Byrd bath is therefore where reconciliation bills are edited in substance. The full mechanics of the rule and its six tests are treated in the series’ dedicated Byrd rule explainer.

In the Byrd bath for the tax bill, the parliamentarian advised that three provisions were extraneous. The first was the short title itself: the clause providing that the act could be cited as the “Tax Cuts and Jobs Act.” A short title has no budgetary effect, which is precisely the problem under the Byrd rule: a provision with no budgetary effect is extraneous to a budget reconciliation bill, and extraneous provisions fall. The second was the Cruz provision expanding section 529 accounts to cover homeschooling expenses, section 11032 of the bill as drafted. The third was part of the excise tax on large university endowments, section 13701, specifically the exception for institutions with fewer than 500 tuition-paying students. None of the three could survive a point of order without sixty votes, and the waiver motion’s 51 to 48 failure confirmed that the votes were not there.

Is the Tax Cuts and Jobs Act the law’s official name?

No. The short title “Tax Cuts and Jobs Act” was struck from the bill on a Byrd-rule point of order on December 19, 2017, before final passage. The enrolled law’s title is the descriptive reconciliation title naming the fiscal year 2018 budget resolution. Everyone still calls it the Tax Cuts and Jobs Act, but no statute bears that name.

The enrolled law’s official title, as recorded in the congressional record, is “An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018.” That is the title on the enrolled bill the president signed. It is descriptive, procedural, and unusable as a citation name, which is why everyone from the Congressional Research Service to the Internal Revenue Service to the press continued to call the law the Tax Cuts and Jobs Act. The continued colloquial use is harmless in conversation and hazardous in citation. A brief that cites “the Tax Cuts and Jobs Act” as the enacted short title misstates the statute’s identity, and the error is revealing: it shows a writer who learned the law from coverage rather than from the enrolled text. The fact that almost no coverage reported the title’s removal is itself part of the story. The most-watched legislative fight of 2017 ended with the bill losing its name on a procedural point, and the name lived on anyway, attached to a law that never bore it.

The attribution for the title’s fall also deserves precision. The parliamentarian did not strike the title alone: the parliamentarian advised, Senator Sanders raised the points of order, Senator Enzi’s waiver motion failed, and the provisions fell. Procedure determined the text, but people operated the procedure.

The reconciliation asymmetry

The single most misunderstood feature of the 2017 act is the difference in durability between its parts. The corporate rate cut to 21 percent has no expiration date. Nearly every individual provision, the rate reductions, the enlarged standard deduction, the cap on the state and local tax deduction, the doubled estate exemption, the pass-through deduction, expires after December 31, 2025. Readers encountering that split naturally assume it reflects a judgment: that Congress considered the corporate cut more important, or more deserving of permanence, than relief for households. That assumption is the central error this profile exists to correct. The split reflects no such judgment. It reflects the Byrd rule.

Why do the individual cuts expire while the corporate rate is permanent?

Reconciliation forbids increasing deficits beyond the ten-year budget window, so the drafters attached sunset dates to the individual provisions to keep the bill’s score inside the $1.5 trillion instruction. The corporate rate carried no sunset because its cost fit the window, a procedural artifact rather than a policy judgment about which provisions deserved to last.

The mechanics are straightforward once stated plainly. The budget resolution allowed a net revenue reduction of no more than $1.5 trillion over fiscal years 2018 through 2027. The Joint Committee on Taxation scored the conference agreement at a $1.456 trillion revenue loss over that window, which fit, but only just. The Byrd rule adds a second, sharper constraint: no provision may increase the deficit in any year beyond the budget window. A permanent extension of the individual provisions would have increased deficits in years after 2027, violating the rule and exposing those provisions to a point of order. The drafters therefore wrote the individual provisions to expire after 2025, so that the official score showed no deficit increase beyond the window. The corporate rate cut, by contrast, was scored as fitting within the permitted totals without a sunset, so it received none. The asymmetry is arithmetic wearing the mask of policy.

This is the reconciliation asymmetry in its full form: the durability of every provision in the 2017 act was determined by a Senate procedural rule rather than by any judgment about which policies deserved to last, and readers who do not know that rule cannot explain the law’s structure. The claim is namable because it is testable. Ask of any provision whether it expires, and the answer will track the scoring constraint, not the provision’s importance. The individual rate cuts were the centerpiece of the bill’s public sale, yet they expire. The corporate alternative minimum tax repeal, a technical item few voters had heard of, is permanent. The difference is not that Congress valued the corporate minimum tax repeal more highly. The difference is that one fit the window and the other did not.

The calibration of the sunsets shows how deliberately the drafters worked the constraint. The Joint Committee’s $1.456 trillion score left $44 billion of room under the $1.5 trillion instruction, a margin of less than 3 percent. That narrow margin is why the conference agreement could not simply make everything permanent and why the final weeks of drafting were dominated by scoring runs rather than policy debates. Each candidate for permanence had to earn its place by fitting inside the remaining room; each provision that did not fit was given a sunset instead. The individual provisions, taken together, were too expensive to make permanent within the instruction, so they were written to expire. The corporate rate cut, priced at 21 percent rather than 20, fit. The sunsets were not scattered randomly across the bill. They were the residual of a subtraction: the instruction minus the permanent provisions, with the individual package sized to the remainder. To call this policymaking is to stretch the term. It was budget engineering, performed under a rule that made budget engineering the only available craft. The drafters could have chosen smaller permanent individual cuts instead of larger temporary ones; they chose the larger temporary version, judging, correctly as it turned out, that the bigger near-term numbers would dominate public perception while the expiration dates would be discounted. The choice was political, but the menu from which it was chosen was procedural.

The Senate rule that did this work deserves its own explanation, and the series provides one at the Byrd rule, explained. The short version relevant here is the “Byrd bath,” the process by which provisions are scrubbed from a reconciliation bill before final passage, described in the section above. The bath is usually described as a technicality, but in December 2017 it deleted the bill’s name, an education-savings expansion, and an endowment-tax carve-out, and it had earlier forced the drafters to write expiration dates into the bill’s most prominent provisions. A procedural screen that can delete a title and rewrite a statute’s lifespan is not a technicality. It is a co-author.

One more clarification belongs in this section, because it is the error the brief flags most urgently. The sunsets are sometimes described as though Congress intended the individual cuts to be temporary on the merits, or as though the expiration dates expressed doubt about the provisions. Nothing in the legislative record supports that reading. The sunset dates were a compliance device, chosen to satisfy the Byrd rule’s out-year test. Whether later Congresses would extend the provisions was a question the 2017 Congress left to the future, and the score was written so that the future would have to act. To read the sunsets as policy is to mistake the cage for the animal. The mechanics of reversion are worth stating precisely: when the individual provisions expire after December 31, 2025, the code reverts to the law as it stood before 2018, with the pre-2018 brackets, standard deduction, exemptions and caps restored, adjusted for the inflation indexing that continued to operate in the background. Reversion is automatic. It requires no vote, no signature and no further action. The 2017 Congress thus legislated twice: once for 2018 through 2025, and once, by inaction, for the years after.

What changed for individuals

The individual provisions were the parts of the statute most households actually felt, and they were also the parts the reconciliation arithmetic treated most harshly. Every major item in this group expires after December 31, 2025. The group is best understood as a package: lower rates, a bigger standard deduction paired with the suspension of the personal exemption, new limits on two itemized deductions, a doubled estate exemption, a doubled child credit, and the zeroing of the health coverage penalty. Taken together they reshaped the individual income tax for eight tax years, and then, by design, they stopped.

The rate structure kept seven brackets but lowered most of them. The act replaced the seven individual brackets of prior law, 10, 15, 25, 28, 33, 35, and 39.6 percent, with a new seven-bracket structure of 10, 12, 22, 24, 32, 35, and 37 percent, applicable to tax years beginning after December 31, 2017 and before January 1, 2026. The top rate fell from 39.6 percent to 37 percent. In 2018, the first year under the new law, the 37 percent bracket began above $500,000 of taxable income for single filers and above $600,000 for joint filers, as announced in the Internal Revenue Service’s inflation-adjustment guidance for that year. The bracket thresholds were also redrawn, so the comparison between old and new rates is not a simple subtraction at every income level, but the direction was uniform: every bracket but the bottom one paid a lower statutory percentage than before. All of it sunsets. After 2025 the brackets revert to the pre-2018 structure unless later legislation says otherwise.

The bracket thresholds, like the standard deduction amounts, were indexed for inflation, but with a change in the inflation measure that compounded over time: the act shifted indexation to the chained consumer price index, which rises more slowly than the traditional consumer price index it replaced. The practical effect is bracket creep in slow motion: because the chained index accounts for consumers substituting away from goods whose prices rise fastest, more income is pushed into higher brackets over time than under the old index, producing a gradual, permanent revenue increase. Unlike the rates and thresholds it adjusts, the chained-index provision has no sunset. It is one of the few individual items the drafters made permanent, presumably because a provision that raises revenue over time helped rather than hurt the score.

The bracket thresholds for joint filers were set at exactly double the single thresholds through much of the rate schedule, which reduced the marriage penalty that had taxed some two-earner couples more heavily than two single individuals with the same combined income. The 2018 top-bracket thresholds, $500,000 for single filers and $600,000 for joint filers, illustrate the pattern. Marriage-penalty relief had been a perennial goal of tax legislation, and the 2017 act achieved more of it than any statute since the early 2000s, though only for the eight years the new brackets applied. Like everything else in the individual group, the relief expires with the brackets that confer it.

The standard deduction was nearly doubled, and the personal exemption was suspended to zero, and the two changes must be read as a pair. For 2018 the standard deduction rose to $12,000 for single filers and married persons filing separately (from $6,350), $18,000 for heads of household (from $9,350), and $24,000 for married couples filing jointly (from $12,700). At the same time the personal exemption, which would have been $4,150 per person in 2018, was set to zero for tax years 2018 through 2025, so a family of four lost $16,600 of exemptions while gaining a larger standard deduction, a trade whose net effect varied by family size and filing situation. Here the near-miss warning matters. Descriptions of the statute sometimes say the personal exemption was repealed. It was not repealed. It was suspended, which is a different legal act with a different consequence: the exemption provision remains in the code, inoperative, and it revives automatically when the suspension lapses after 2025. Repeal would have required affirmative legislation to restore it. Suspension restored it by the calendar.

What did the SALT cap actually limit?

The cap limited the itemized deduction for state and local taxes to $10,000 per return ($5,000 for married filing separately), covering income or sales taxes plus property taxes not tied to a trade or business. It did not touch business state and local taxes, which remained fully deductible, and it expired on the same individual-provision schedule.

The state and local tax deduction had been part of the code since the modern income tax began, and capping it at $10,000 was among the most contested changes in the statute. The cap applies to the aggregate of state and local income taxes (or sales taxes, at the filer’s election) and property taxes, but only to the portion not incurred in carrying on a trade or business. Business state and local taxes remained fully deductible as ordinary business expenses, a distinction that matters enormously to the provision’s real incidence. For an employee in a high-tax state who itemized, the cap raised taxable income substantially. For a business owner deducting the same taxes as business expenses, nothing changed. Before the cap, a taxpayer who itemized could deduct the full amount of state income and property taxes paid, which meant the federal government effectively subsidized a portion of high state and local tax burdens. The cap ended that open-ended subsidy for amounts above $10,000. The cap was also one of the provisions most visibly shaped by the scoring constraint: a permanent cap would have raised revenue permanently, which would have helped the score, but the cap was bundled with the expiring individual package and given the same sunset. It applies to tax years 2018 through 2025.

One interaction softened the cap’s bite for some households: state and local taxes had never been deductible under the individual alternative minimum tax, so households that paid the AMT under prior law had already been losing the SALT deduction in practice. Tax provisions are rarely experienced in isolation, and the SALT cap is a case study in why the code must be read as a system rather than a list.

The interaction between the SALT cap and the doubled standard deduction reshaped itemizing itself. With the standard deduction at $24,000 for joint filers and the SALT deduction capped at $10,000, millions of taxpayers who had itemized under prior law found that the standard deduction now exceeded their total itemized deductions, and they stopped itemizing. Analysts estimated that the share of filers itemizing fell from roughly 30 percent to roughly 10 percent, one of the largest single-year changes in filing behavior the code had ever produced. That shift had second-order consequences far beyond the SALT cap: the charitable deduction, which is only available to itemizers, effectively disappeared for most middle-income donors, changing the economics of charitable giving without any direct change to the charitable deduction’s rules. The mortgage interest deduction, likewise available only to itemizers, lost relevance for millions of homeowners. A statute that never touched the charitable deduction’s text nonetheless transformed its reach, which is a useful reminder that tax provisions operate as a system, and that changing the standard deduction changes every deduction conditioned on itemizing.

The state-level response to the cap became a minor federalism drama of its own. Several high-tax states sued the federal government, arguing the cap unconstitutionally intruded on state taxing authority; the Southern District of New York dismissed the suit, and the Second Circuit affirmed, ending that challenge. A more durable response came in the form of pass-through entity taxes, under which states imposed entity-level taxes on partnerships and S corporations, deductible at the entity level outside the SALT cap, and gave owners offsetting credits. By the early 2020s a large majority of states had enacted some version of the entity-level tax, narrowing the cap’s bite without removing it.

The mortgage interest deduction was narrowed rather than eliminated. The act lowered the cap on the mortgage interest deduction for new debt to $750,000 of acquisition indebtedness, applicable to debt incurred after December 15, 2017. Debt incurred on or before that date was grandfathered under the prior $1,000,000 cap. The December 15 date is worth fixing because it created a two-week transition cohort: taxpayers who closed before mid-December 2017 kept the old cap, while those who closed after faced the new one. The act also eliminated the deduction for interest on home equity lines of credit for tax years 2018 through 2025, regardless of when the line was established, a change that surprised many homeowners who had treated home equity borrowing as permanently deductible. The $500,000 cap that appears in some summaries was the House bill’s figure; the enacted cap is $750,000, and the distinction is another of the draft-versus-law errors that careful readers learn to check. The grandfathering rule created a two-tier system for the years the provision applied: neighbors with identical homes could face different deduction ceilings based solely on when they borrowed.

The estate tax exemption was doubled. The basic exclusion amount rose to $11,180,000 per decedent for 2018, up from $5,490,000 in 2017, with the amount indexed for inflation in subsequent years. In practical terms, the change removed all but the very largest estates from the federal estate tax for the duration of the provision, which sunsets after 2025. The doubling was achieved by statute rather than by regulation, and like the other individual provisions it carries the reconciliation fingerprint: a permanent doubling would have added to deficits beyond the budget window, so the statute doubled the exemption for eight years instead. Estate planners spent the window advising clients on the use-it-or-lose-it character of the doubled exemption, including the question whether gifts made under the doubled exemption would be clawed back if the exemption later fell, a question the Treasury Department answered with regulations providing no clawback.

The individual alternative minimum tax survived, but in weakened form. The act retained the individual AMT while raising its exemption amounts, to $70,300 for single filers and $109,400 for joint filers in 2018, and raising the phaseout thresholds to $500,000 and $1,000,000 respectively. The higher thresholds removed millions of taxpayers from the AMT’s reach for the duration, another individual provision scheduled to expire after 2025. The corporate AMT, by contrast, was repealed outright and permanently, a contrast that again displays the asymmetry: two minimum taxes, one for individuals and one for corporations, treated oppositely on permanence because the budget arithmetic, not any theory of minimum taxation, dictated the outcome.

The child tax credit was doubled to $2,000 per qualifying child, with up to $1,400 refundable, and its phaseout thresholds were raised substantially, extending the credit well into the upper middle class. The credit expansion was one of the provisions most directly felt by families, and it was negotiated in the Senate as the price of support from members who wanted the bill’s benefits tilted further toward households with children. Like the rest of the individual package, it expires after 2025.

The final individual provision to note is the one with the longest shadow: section 11081 of the act reduced the Affordable Care Act’s individual shared responsibility payment, the penalty for lacking health coverage, to zero for tax year 2019 and all subsequent years. The mandate’s statutory language was not repealed. Only the penalty was zeroed, and the distinction carried constitutional consequences that played out in later litigation over whether the mandate, once unenforceable, could still be severed from the rest of the health law. Within the tax story, the zeroing matters for two reasons. First, it was scored as a revenue and outlay provision: the Congressional Budget Office estimated that eliminating the penalty would reduce federal outlays for health insurance subsidies, because fewer people would enroll in subsidized coverage, and that outlay reduction helped the bill fit inside the reconciliation cap. A health provision did budget work inside a tax bill. Second, the zeroing was made permanent. Unlike the rate cuts and the SALT cap, the mandate penalty reduction carried no sunset, because reducing outlays does not increase deficits, and the Byrd rule’s deficit test therefore gave it a pass. The provision that looked least like a tax cut was the individual provision most securely permanent, which is exactly the kind of inversion the reconciliation asymmetry produces. The later history of the zeroed penalty, including the litigation it generated, is treated in Affordable Care Act: later changes.

Stepping back, the individual package had a coherent shape despite its procedural origins. It moved the code toward fewer itemizers and concentrated its largest dollar benefits on higher-income households, a distributional fact the evidence section will quantify. And it was, in its entirety, temporary. Every rate, every threshold, every cap in this section carried the same expiration date, because the Byrd rule’s out-year test applied to all of them equally. The individual provisions were the bill’s public face and its procedural sacrifice at once.

What changed for businesses

If the individual provisions were the bill’s public face, the business provisions were its structural core, and they were built to last. The corporate rate cut has no sunset. The corporate alternative minimum tax was repealed outright. The international rules described in the next section were written as permanent regimes. The business group is where the reconciliation asymmetry shows its other face: the provisions the drafters could afford to make permanent, within the scoring window, were made permanent, and the result is a business tax code that changed fundamentally in 2017 while the individual code changed only until 2025.

Why did the corporate rate land at 21 percent?

Both chambers’ pre-conference bills set the rate at 20 percent, but the conference agreement raised it to 21 percent to make the ten-year revenue score fit the reconciliation instruction. The flat 21 percent rate applies to taxable years beginning after December 31, 2017, and unlike the individual provisions it carries no sunset date.

The corporate rate cut was the largest single item in the statute by revenue cost. The prior rate was a graduated structure topping out at 35 percent; the statute replaced it with a flat 21 percent applying to all corporate taxable income for tax years beginning after December 31, 2017. The choice of 21 rather than 20 is a small window into how reconciliation legislating works: the conference committee needed the final score to fit the $1.5 trillion instruction, and the extra point of rate bought the needed room. The 20 percent figure that appears in some early accounts was real, but it belonged to the pre-conference bills, not to the enacted law. Any citation of a 20 percent enacted rate repeats a drafting-stage number as though it were law. The new rate also repositioned the United States in the international rate rankings: the combined federal-state statutory rate fell from the highest in the developed world to roughly the middle of the pack, which was precisely the competitive repositioning the drafters sought. Whether the repositioning produced the investment and wage gains its advocates projected is the question the evidence sections answer; that it repositioned the rate is the undisputed fact. A single percentage point on the corporate base was worth roughly $100 billion over the window, which made it the cheapest large lever the conferees had.

The corporate alternative minimum tax was repealed in full. The corporate AMT had operated as a parallel system ensuring that corporations with substantial economic income paid at least some tax; the statute eliminated it for tax years beginning after December 31, 2017, with provisions allowing corporations to use accumulated AMT credits against regular liability and to claim refunds of remaining credits over several years. The repeal is permanent, its permanence reflecting the scoring: repealing the corporate AMT cost revenue within the window, but the drafters judged the cost fittable, and no sunset was attached.

Bonus depreciation let firms expense new investment immediately, on a schedule that phases down over time. For qualified property acquired and placed in service after September 27, 2017 and before January 1, 2023, the statute allowed a 100 percent first-year deduction. The percentage then steps down under the enacted schedule: 80 percent for property placed in service in 2023, 60 percent in 2024, 40 percent in 2025, 20 percent in 2026, and zero in 2027, with long-production-period property on a one-year-delayed schedule. The phase-down was itself a scoring device, spreading the provision’s cost so the bill fit the window. Two features of the provision deserve note. First, the statute extended bonus depreciation to used property, where prior law had limited it to new investment, which broadened the incentive to acquisitions of existing assets. Second, the September 27, 2017 acquisition date meant the 100 percent rate applied retroactively to property acquired after that date but before enactment, rewarding investment made in anticipation of the bill. Unlike the individual sunsets, which were Byrd-rule compliance devices, the phase-down was a designed cost-control schedule: it let the writers claim the full incentive effect up front while booking a declining cost over the budget window. Economists debate whether scheduled phase-downs actually influence investment timing or merely shift it, since firms anticipating the step-down have an incentive to accelerate purchases into the 100 percent years, pulling investment forward rather than increasing its total.

The new limit on business interest deductions, codified at section 163(j), capped the amount of business interest a firm could deduct each year at the sum of its business interest income, 30 percent of its adjusted taxable income, and its floor-plan financing interest. Interest disallowed under the cap carries forward indefinitely to future years rather than being lost. Small businesses, defined by average gross receipts of $25 million or less over three years, inflation-adjusted, are exempt. The provision is permanent, though its bite was scheduled to sharpen: adjusted taxable income was computed on an EBITDA-like basis before 2022 and on an EBIT-like basis starting in 2022, which lowered the 30 percent ceiling and disallowed more interest. The limit was the statute’s main base-broadener on the business side, partially offsetting the revenue cost of the rate cut, and its design shows the drafters reaching for permanent revenue raisers to pay, within the window, for permanent rate relief. It was designed as permanent base protection paired with the lower corporate rate, ensuring that heavily indebted firms could not deduct their way to near-zero liability under the new rate.

How does the pass-through deduction work?

Section 199A lets owners of pass-through businesses deduct up to 20 percent of qualified business income below the line, without itemizing. Wage and capital limits and exclusions for specified service businesses phase in above income thresholds. The deduction sunsets after 2025 with the rest of the individual provisions, since it was scored as an individual tax cut.

Section 199A created a new deduction for qualified business income from pass-through entities: partnerships, S corporations and sole proprietorships. The deduction equals 20 percent of qualified business income, but three limits shape it. First, it is capped at 20 percent of the owner’s taxable income, computed before the deduction and excluding net capital gains, so it cannot create or enlarge a loss. Second, above income thresholds the deduction is limited to the greater of 50 percent of W-2 wages paid by the business or 25 percent of wages plus 2.5 percent of the unadjusted basis of qualified property, a wage-and-capital test that favors businesses with payrolls and tangible investment over businesses that are essentially their owners’ labor. Third, above the same thresholds, income from specified service trades or businesses, a statutory list covering health, law, accounting, actuarial science, performing arts, consulting, athletics, financial services, brokerage services and any business whose principal asset is the reputation or skill of its employees, is excluded entirely. Engineers and architects were carved out of the service-business exclusion after lobbying that their work was capital-intensive rather than purely personal-service. The three limits phase in over an income range rather than applying all at once, which made the provision one of the most complex items in the statute and a rich source of planning and controversy.

The deduction was the statute’s answer to an obvious question: if corporations received a 21 percent rate, what relief would the millions of businesses taxed through their owners’ individual returns receive? The answer was a 20 percent deduction rather than a rate cut, and unlike the corporate cut it was given the individual sunset. One design choice softened the provision’s complexity for ordinary filers: because the deduction is claimed below the line, in computing taxable income rather than adjusted gross income, owners who take the standard deduction can still claim it, which means the SALT cap and the standard-deduction increase did not indirectly claw it back. Section 199A expires after 2025 with the rest of the individual package, which means the pass-through sector’s relief was temporary while the corporate sector’s was permanent. The asymmetry is procedural, but its economic consequence is real: two businesses with identical income can face very different long-run tax treatment depending solely on their legal form, because one form’s relief was scored inside the window and the other’s was not.

Read as a whole, the business group moved the corporate tax from a high-rate, base-narrowing system toward a lower-rate, broader-base one, while leaving the pass-through sector with a large but temporary deduction. The permanence of the corporate changes is what makes the 2017 act a structural break rather than a temporary stimulus. The statute’s answer to the old graduated 35 percent system was to cut the headline rate nearly in half and pay for part of the cut by broadening the base: limiting interest deductions, repealing the corporate alternative minimum tax’s parallel system, and imposing the new international minimum taxes described in the next section. Rate down, base broader, system simpler in structure if not in detail. That was the design logic, and the permanence of the corporate items is what let the logic endure beyond the budget window.

What changed across borders

The international provisions were the least understood part of the statute in public debate and the most technically ambitious part of the drafting. Before 2017 the United States taxed the worldwide income of its resident corporations, with tax on foreign earnings generally deferred until those earnings were brought home. The result was an enormous stockpile of untaxed foreign earnings held abroad and a steady incentive to book profits in low-tax jurisdictions. The statute moved the system toward a territorial model, in which foreign earnings are largely exempt, while surrounding that exemption with new anti-abuse regimes. The territorial shift was permanent. The regimes policing it were permanent too, though their rates were scheduled to change.

The old system the statute replaced is worth describing precisely, because the new system’s logic is a response to its failures. Under pre-2018 law, the United States asserted taxing jurisdiction over the worldwide income of domestic corporations, but tax on the active earnings of foreign subsidiaries was deferred until those earnings were distributed as dividends to the American parent. Since bringing the money home triggered the tax, firms did not bring it home, and the stockpile of indefinitely reinvested foreign earnings grew into the trillions of dollars. The system also encouraged inversions, in which an American firm would merge with a smaller foreign firm and reincorporate abroad to escape the worldwide net entirely. The worldwide-with-deferral system taxed foreign earnings at 35 percent in theory and at zero in practice for any firm willing to wait, which was nearly all of them.

The participation exemption, codified at section 245A, is the territorial core. It allows a domestic corporation a 100 percent dividends-received deduction for the foreign-source portion of dividends received from 10-percent-owned foreign corporations, subject to a one-year holding period and excluding dividends from regulated investment companies, real estate investment trusts and hybrid arrangements. In plain terms, a U.S. multinational that earns profits through a foreign subsidiary and brings those profits home as dividends generally owes no U.S. tax on them. That is the territorial principle: the United States taxes domestic income and largely exempts foreign active business income. The exemption applies to distributions after December 31, 2017, and it carries no sunset.

What did the transition tax do to old foreign profits?

Section 965 imposed a one-time tax on accumulated post-1986 foreign earnings, at 15.5 percent on cash holdings and 8 percent on the rest, payable in eight installments. It cleared the old offshore stockpile so the new participation exemption could start from zero. Research links the resulting repatriations more to share repurchases than to new domestic investment.

Because the old worldwide system had left years of foreign earnings untaxed and sitting offshore, the statute imposed a one-time transition tax to clear the books. Section 965 deemed the accumulated post-1986 foreign earnings of specified foreign corporations repatriated in the last tax year beginning before January 1, 2018, taxing cash and cash-equivalent holdings at 15.5 percent and the remainder at 8 percent, with the liability payable in eight annual installments. The rates were achieved through a special deduction under section 965(c), which means the effective rates for non-corporate taxpayers could differ slightly from the headline figures. The measurement date was the greater of the earnings measured on November 2, 2017 or on December 31, 2017. The transition tax was one-time by design rather than by sunset: it applied to a fixed stock of past earnings and then ended, because there was nothing left to tax. It raised substantial revenue in the scoring window, which helped pay for the permanent rate cut, and it ended the deferral era in a single stroke. The two-rate structure reflected a practical judgment: cash could be taxed more heavily because it was liquid and available to pay the tax, while illiquid earnings, tied up in factories and equipment abroad, were taxed more lightly to avoid forcing distress sales. The eight-installment payment schedule, back-loaded toward the later years, spread the collection across the budget window in a pattern the estimators could score with confidence.

The statute’s answer had three moving parts, and they must be understood as a system rather than a list. The participation exemption removed the tax on repatriated dividends, which ended the lockout effect at a stroke: there was no longer any tax reason to keep earnings abroad, because bringing them home cost nothing. The transition tax settled the past, imposing a one-time charge on the accumulated earnings that the old system had left untaxed, so the move to territoriality did not simply forgive the stockpile. And the three new regimes, GILTI, FDII and BEAT, policed the future, taxing the categories of income most easily shifted across borders. Exemption without the backstops would have been an invitation to shift profits offshore and bring them home tax-free. The backstops without the exemption would have preserved the lockout. Together they formed a coherent alternative to the old system, and their coherence is the strongest defense of the international title against the charge that it was merely a giveaway.

The new regimes that police the territorial system are where the statute’s acronyms multiply. GILTI, the tax on global intangible low-taxed income under section 951A, imposes current-year U.S. tax on a U.S. shareholder’s net controlled-foreign-corporation income exceeding a 10 percent return on tangible depreciable assets, the theory being that returns above that threshold reflect mobile intangible income. A 50 percent deduction under section 250 produced an effective corporate rate of 10.5 percent on GILTI as enacted. FDII, the deduction for foreign-derived intangible income under section 250, works the other direction: it gives domestic corporations a 37.5 percent deduction on income from selling goods and services abroad that exceeds a normal return on tangible assets, producing an effective rate of 13.125 percent as enacted, as an incentive to locate intangible income in the United States rather than offshore. The two were designed as a pair, a stick for intangible income booked abroad and a carrot for intangible income booked at home. Both are permanent regimes, but their rates were scheduled to rise after 2025 as part of the long-run scoring, another instance of the calendar doing the work of policy.

BEAT, the base erosion and anti-abuse tax under section 59A, is the backstop. It imposes a minimum tax on large multinational groups that strip the U.S. tax base through deductible payments to foreign affiliates, such as interest, royalties and service fees. It applies to taxpayers with average annual gross receipts of $500 million or more over three years whose base-erosion percentage reaches 3 percent, or 2 percent for banks and registered securities dealers. The enacted rate schedule was 5 percent for 2018, 10 percent for 2019 through 2025, and 12.5 percent thereafter, with each rate one point higher for banks and registered securities dealers. The tax works by recomputing the taxpayer’s liability without the benefit of the outbound deductible payments and collecting the excess of that recomputed amount over the regular tax, which means it binds only when the firm’s deductible outbound payments are large relative to its income. It was the bluntest instrument in the international title, and its bluntness drew criticism from trading partners and from domestic firms whose ordinary cross-border payments, for inventory or services, were swept into a regime aimed at profit shifting. Firms responded by restructuring their payment flows, and the provision’s complexity made it one of the most heavily lobbied parts of the international title during the regulatory phase. BEAT was the statute’s answer to the oldest criticism of territorial systems: that exempting foreign income invites firms to shift domestic income offshore.

The mechanics of the backstops show how the drafters translated economic concepts into statutory formulas. GILTI starts from a normal return: the statute deems 10 percent of a controlled foreign corporation’s tangible depreciable assets to be a routine return on physical investment, and taxes only the income above that threshold as presumptively mobile intangible income, with the 50 percent deduction halving the effective rate to 10.5 percent. FDII mirrors the logic domestically: income from foreign sales exceeding the same kind of routine return gets a 37.5 percent deduction, for an effective 13.125 percent rate, so that serving foreign markets from the United States is taxed more lightly than serving them from a haven but less lightly than under the old 35 percent regime. BEAT operates differently, as a minimum tax rather than a rate on a category: it recomputes the tax of large multinationals without the benefit of deductible payments to foreign affiliates and collects the difference if that recomputed tax exceeds the regular liability. Each formula embodies a judgment about where avoidance concentrates, and each judgment was, in the end, a guess tested by experience.

The international group, taken together, replaced deferral with exemption-plus-backstops. The old system had taxed foreign earnings eventually and encouraged firms to keep them abroad indefinitely. The new system exempts them and taxes the margins where avoidance concentrates: intangible income, deductible outbound payments, and the one-time stock of past earnings. The design was permanent because the drafters could score it permanently within the window, and the rate schedules that step up after 2025 show the same scoring logic at work in miniature.

Implementation: the statute meets the world

A statute of this size is never finished at enactment. The 2017 act delegated enormous interpretive work to the Treasury Department and the Internal Revenue Service, which spent the next three years issuing proposed and final regulations, notices and revenue procedures across nearly every title of the bill. The volume of guidance was itself a measure of the statute’s complexity: section 199A alone required regulations defining the boundaries of specified service trades or businesses, de minimis exceptions for firms with small amounts of service income, and rules for aggregating related businesses for the wage-and-capital limits. The section 199A regulations, finalized in early 2019, ran to hundreds of pages and created safe harbors, such as a rental real estate safe harbor for landlords seeking to qualify their rental income, that effectively extended the statute by administrative action. Practitioners spent the first two years of the law’s operation waiting for answers to questions the statutory text had left open, from the treatment of tiered partnerships to the definition of reputation-or-skill businesses, and the answers, when they came, reshaped planning in ways the drafters had not fully anticipated. The episode is a reminder that a complex statute’s real content is the text plus the regulations, and that the regulations are written under time pressure by an agency interpreting language it did not draft. Section 163(j) required guidance on the computation of adjusted taxable income and the treatment of partnerships. The international provisions required guidance on the GILTI high-tax exceptions, the FDII documentation rules, the BEAT’s application to specific payment types, and the transition tax’s installment mechanics. Each regulation narrowed or clarified the statute, and each became, for practical purposes, part of the law as taxpayers experienced it. A profile of the enacted text cannot substitute for the regulatory history, but the reader should know the text was only the beginning.

The SALT cap provoked the most visible implementation fight, and the Treasury’s response to the state workarounds became its own body of law. When states created charitable funds offering state tax credits for contributions, attempting to convert capped SALT payments into uncapped charitable deductions, the Treasury answered with final regulations requiring taxpayers to reduce their charitable contribution deduction by the amount of any state or local tax credit received, applying the longstanding quid pro quo principle to shut the workaround down. The more durable state response, entity-level taxes on pass-through businesses deductible at the entity level and therefore outside the individual cap, was blessed by the IRS in Notice 2020-75, and a large majority of states enacted such pass-through entity taxes in the years that followed. The episode illustrates a general truth about the statute: a cap written for individuals could be routed around by entities, because the code taxes the two differently and the statute preserved the difference.

The mandate zeroing produced its own aftershock in the courts. With the penalty at zero, challengers argued the mandate could no longer be sustained as an exercise of the taxing power and that the rest of the Affordable Care Act must fall with it. The Supreme Court resolved the challenge in 2021 on standing grounds, by 7 to 2, without reaching the constitutional merits, leaving the zeroed penalty and the surviving mandate language in place. The litigation is part of the Affordable Care Act’s later history rather than this statute’s, but its origin was section 11081 of the 2017 act, another instance of a scoring-driven provision generating consequences its drafters had not centered.

The transition tax’s collection mechanics showed the statute’s practical side. The liability, though deemed to arise in the 2017 tax year, was payable in eight annual installments, back-loaded so that the largest payments came latest. The installment schedule was a political and practical necessity: demanding immediate payment of tax on decades of accumulated earnings would have strained even the largest firms’ cash positions, and the phase-in gave the revenue estimators a smooth collection path to score. Like much of the statute, the schedule was designed as much for the scorekeepers as for the taxpayers.

The permanent-or-expiring table

Each major provision appears below with its effective date, its durability, and the reason for that durability. Read down the third column and the reconciliation asymmetry becomes visible as a pattern rather than an argument: the individual provisions expire because the Byrd rule required it, the business and international provisions survive because the budget window permitted it, and the few provisions with their own phase-down schedules carry the cost-control logic the writers used to fit the $1.5 trillion envelope.

Provision	Effective date	Permanent or scheduled to expire	Reason for that treatment
Individual rate reductions (10, 12, 22, 24, 32, 35, 37 percent)	Tax years beginning after December 31, 2017	Scheduled to expire after December 31, 2025	Byrd-rule compliance: permanent cuts would increase deficits beyond the budget window
Standard deduction increase and personal exemption suspension	Tax years 2018 through 2025	Scheduled to expire after December 31, 2025	Byrd-rule compliance: same deficit-beyond-window bar
SALT deduction cap ($10,000; $5,000 married filing separately)	Tax years 2018 through 2025	Scheduled to expire after December 31, 2025	Byrd-rule compliance: the cap’s revenue partially offset the rate cuts inside the window
Mortgage interest cap ($750,000, debt after December 15, 2017)	Tax years 2018 through 2025	Scheduled to expire after December 31, 2025	Byrd-rule compliance: individual-side provision under the same bar
Estate tax exemption doubling ($11,180,000 in 2018)	Tax years 2018 through 2025	Scheduled to expire after December 31, 2025	Byrd-rule compliance: permanent doubling would breach the beyond-window deficit test
Section 199A pass-through deduction (20 percent of qualified business income)	Tax years 2018 through 2025	Scheduled to expire after December 31, 2025	Scored as an individual provision, so it rides the individual sunset
Bonus depreciation (100 percent, then phase-down)	Property placed in service after September 27, 2017	Phase-down: 80 percent (2023), 60 percent (2024), 40 percent (2025), 20 percent (2026), zero (2027)	Statutory cost-control schedule, not a Byrd sunset; designed to taper the incentive
Section 163(j) business interest limit (30 percent of adjusted taxable income)	Tax years beginning after December 31, 2017	Permanent	Base-protection measure whose cost fit inside the window; no sunset written
Corporate rate reduction (35 percent to flat 21 percent)	Tax years beginning after December 31, 2017	Permanent	Fit within the $1.5 trillion reconciliation allowance over the window
Corporate alternative minimum tax repeal	Tax years beginning after December 31, 2017	Permanent	Same window arithmetic as the rate cut
Section 245A participation exemption (100 percent dividends-received deduction)	Distributions after December 31, 2017	Permanent	Territorial architecture designed as the lasting international system
GILTI (10.5 percent effective rate via 50 percent section 250 deduction)	Tax years beginning after December 31, 2017	Permanent, with the deduction scheduled to narrow after 2025	Backstop to the participation exemption; rate tightening manages cost
FDII (13.125 percent effective rate via 37.5 percent deduction)	Tax years beginning after December 31, 2017	Permanent, with the deduction scheduled to narrow after 2025	Export incentive paired with GILTI; same scheduled tightening
BEAT (5 percent for 2018, 10 percent for 2019 through 2025, 12.5 percent after)	Tax years beginning after December 31, 2017	Permanent, with the scheduled rate increases	Anti-abuse minimum tax; the rising schedule is the cost control
Section 965 transition tax (15.5 percent on cash, 8 percent on the rest)	Deemed repatriation in the last tax year beginning before January 1, 2018	One-time, payable in eight installments	Cleared the pre-2018 offshore stockpile so the territorial system could start clean
Individual shared responsibility payment reduced to zero	Tax years beginning after December 31, 2018	Permanent	Reducing the penalty lowers outlays rather than raising deficits, so the Byrd test passed it

Two patterns in the table deserve underlining. First, the “Reason for that treatment” column contains only three distinct logics: the Byrd rule’s beyond-window deficit bar, the reconciliation envelope’s within-window arithmetic, and designed phase-down schedules. No provision’s durability reflects a freestanding judgment that it deserved permanence. Second, the table exposes the provisions whose treatment surprises newcomers: the mandate zeroing, permanent despite being an individual-side change, because it saved rather than cost money; the 199A deduction, expiring despite being a business provision, because it was scored on the individual side; and the international regimes, permanent despite being the most experimental part of the law, because their costs fit the window. Each surprise dissolves once the procedural logic is visible, which is why the table belongs at the center of any serious reading of the statute.

Two cautions accompany the table. First, “permanent” in this context means the provision was enacted without a sunset date, not that later Congresses cannot change it. Congress revisits the code continually. Second, the table describes the statute as enacted in December 2017: the expiration dates and rate schedules are legal facts about the enrolled text, and later Congresses have since modified some of these provisions, including bonus depreciation and the post-2025 international rate schedules.

A worked example: one family’s tax bill, old law vs. new

Abstraction is the enemy of understanding in tax legislation, so consider a concrete household: a married couple filing jointly with two qualifying children and $120,000 of wage income, taking the standard deduction. Under 2017 law, their standard deduction was $12,700 and their personal exemptions were $4,050 per person, $16,200 for four, leaving taxable income of about $91,100. Applying the 2017 brackets, 10 percent to $18,650, 15 percent to $75,900, and 25 percent above, their tax before credits was roughly $14,250. The child tax credit was $1,000 per child, but it phased out above $110,000 of income for joint filers, so at $120,000 they lost $500 of the $2,000 total, keeping $1,500. Their net federal income tax was roughly $12,750.

Under the 2018 law, the same family’s standard deduction was $24,000 and their personal exemptions were zero, leaving taxable income of $96,000. Note the paradox: their taxable income rose, because the lost $16,200 of exemptions exceeded the $11,300 gain in the standard deduction. But the new brackets, 10 percent to $19,050, 12 percent to $77,400, and 22 percent above, applied lower rates to that larger base, producing tax before credits of roughly $13,000. The child tax credit was now $2,000 per child with phaseout beginning at $400,000, so they kept the full $4,000. Their net tax was roughly $9,000, a saving of about $3,750, or nearly 30 percent. The example shows the act’s individual package working as designed for a middle-class family: the doubled credit did more work than the rate cuts, the larger standard deduction more than offset the lost exemptions once credits were counted, and the SALT cap was irrelevant because the family did not itemize.

Now change one fact: give the family $400,000 of income, a $14,000 property tax bill, and $20,000 of state income tax, and have them itemize. Under old law they deducted the full $34,000 of state and local taxes; under the new law they deduct $10,000. At a 32 or 35 percent marginal rate, the lost $24,000 of deductions costs them roughly $8,000 of additional tax, offsetting much of what the lower rates saved them. This is the SALT cap’s concentrated footprint in miniature, and it explains why the act’s distributional tables show broad average gains alongside intense geographic concentration of pain.

The quiet revenue raisers inside a tax cut

A $1.5 trillion net tax cut still contained provisions that raised revenue, because the envelope required offsets and the writers preferred raising revenue quietly to cutting tax cuts loudly. The most durable of the quiet raisers was the switch to chained CPI-U for inflation indexing. The chained index grows more slowly than the traditional consumer price index, roughly a quarter point per year, so bracket thresholds and other indexed parameters rise more slowly, pushing taxpayers into higher brackets faster over time. The effect compounds, and because it applies to the permanent provisions as well as the expiring ones, it keeps raising revenue after the individual sunsets take effect.

The SALT cap was the largest explicit base-broadener on the individual side, raising hundreds of billions over the window by ending the unlimited deduction for state and local taxes above $10,000. The mortgage interest cap’s reduction to $750,000 for new debt raised a smaller amount, as did the elimination of the home equity interest deduction. On the business side, the section 163(j) interest limitation raised revenue by capping a deduction that had been unlimited, and its scheduled tightening, from the EBITDA-based measure of adjusted taxable income to the EBIT-based measure starting in 2022, was itself a phased revenue raiser. The international title’s scheduled narrowings, the GILTI and FDII deductions shrinking after 2025 and the BEAT rate rising to 12.5 percent, booked additional window revenue in the later years. And the individual sunsets were the largest raiser of all in a sense: by ending the individual provisions after 2025, the bill avoided booking two years of their cost, a timing choice worth hundreds of billions within the window.

The mandate zeroing worked the revenue logic in reverse. Reducing the shared responsibility payment to zero was scored as reducing federal outlays, because the Congressional Budget Office projected that fewer people would enroll in subsidized coverage without the penalty, and lower enrollment meant lower subsidy spending. That outlay reduction improved the bill’s net score, effectively paying for a portion of the tax cuts with health-care savings. The episode is a reminder that reconciliation scorekeeping treats the federal budget as a single ledger: a health provision can finance a tax cut, and the Byrd rule’s deficit tests apply to the net, not to each title separately.

Why the effective dates matter

One of the statute’s quiet traps is its variety of effective dates, and readers who assume everything started on January 1, 2018 will misread several provisions. The general rule for the individual, business, and international titles was tax years beginning after December 31, 2017, which for calendar-year taxpayers meant January 1, 2018, but for fiscal-year taxpayers meant the start of their first fiscal year after that date. The distinction mattered for corporations with non-calendar years, whose 21 percent rate and new international regimes phased in on their own fiscal calendars rather than the calendar year.

Three provisions deliberately broke the general pattern. Bonus depreciation reached back to September 27, 2017, the date the House leadership released its tax framework, so property acquired and placed in service in the last quarter of 2017 qualified for 100 percent expensing. The retroactivity rewarded investment made in anticipation of the law and created a short window in which taxpayers who had been waiting for clarity could act with the benefit of hindsight. The mortgage interest cap used December 15, 2017 as its dividing line: debt incurred on or before that date kept the $1,000,000 cap, while debt incurred after faced the $750,000 limit. The two-week transition cohort, homebuyers who closed in mid-December, lived under different rules depending on which side of the date their closing fell, a vivid example of how effective dates create arbitrary winners and losers.

The mandate zeroing used yet another date: tax years beginning after December 31, 2018, meaning the penalty fell to zero starting with the 2019 tax year, a full year after the rest of the individual provisions took effect. The delay was a scoring choice, pushing the outlay effects one year later, and it meant that the 2018 tax year operated under the old penalty while everything else had changed. The section 965 transition tax looked backward rather than forward, measuring accumulated earnings by the greater of the November 2, 2017 or December 31, 2017 figures, dates chosen to prevent taxpayers from manipulating their cash positions between announcement and enactment. Each of these dates tells a small story about legislative strategy: retroactivity to reward anticipation, grandfathering to protect reliance, delay to manage scoring, and measurement dates to prevent gaming. A reader who checks the effective date of each provision, as the permanent-or-expiring table does, will avoid the most common mechanical errors in applying the statute.

How to read the score: a note on JCT conventions

The revenue numbers in the next section come with conventions that determine what they mean, and a reader who does not know the conventions will misread the numbers. The Joint Committee on Taxation produces two kinds of estimates. The conventional estimate, the one that governed the reconciliation process, incorporates taxpayers’ behavioral responses to changed incentives: income shifting across years, changes in business organization, realizations of capital gains timed to rate changes. It does not incorporate macroeconomic feedback: changes in overall output, employment, or growth resulting from the bill. That exclusion is not an oversight. It is the convention, adopted because macroeconomic feedback is harder to estimate reliably and more sensitive to modeling assumptions, and because the budget process needs a stable, agreed-upon measuring rod. The bill’s supporters argued throughout the debate that the convention understated the bill’s fiscal responsibility by ignoring growth; the scorekeepers replied that their job was measurement under fixed rules, not forecasting under the most favorable assumptions.

The dynamic estimate, which the Joint Committee and the Congressional Budget Office produce separately, adds macroeconomic feedback through models of how tax changes affect labor supply, investment, and output. The Penn Wharton Budget Model’s dynamic estimate for the conference agreement projected 0.6 to 1.1 percent higher gross domestic product by 2027, offsetting a fraction of the static revenue loss. The range itself was the message: dynamic scoring does not produce a single answer but a band of answers depending on assumptions about how responsive the economy is to tax changes, and the band for this bill ran from modest to moderate, never approaching self-financing. Readers encountering dynamic estimates should always ask which model’s assumptions produced them, because the Tax Foundation’s model, with different assumptions about investment responsiveness, produced larger gains, and the choice of model was doing as much work as the choice of policy.

Two more conventions matter. First, the estimates are net of outlay effects: the $1.456 trillion figure includes both revenue reductions and outlay changes, which is why the mandate zeroing’s projected reduction in health subsidy spending improved the score. A reader who thinks of the number as purely a tax-revenue figure will misunderstand what it contains. Second, the window is everything: the estimate covers fiscal years 2018 through 2027, and provisions whose costs fall mostly outside the window look cheaper than provisions whose costs fall inside it. The individual sunsets were the purest exploitation of this convention, booking eight years of cost for provisions whose policy was open-ended. None of this implies the conventions are wrong; they are the agreed rules of the budget game, applied to every reconciliation bill. It implies only that the numbers are measurements under rules, not photographs of the future, and that the rules themselves shaped the bill the numbers were measuring.

What the estimators said it would cost

Revenue estimates are the least read and most load-bearing part of any tax statute’s story. The 2017 act’s estimates deserve close attention because they were both the design constraint and the first draft of the verdict. The official estimator is the Joint Committee on Taxation, a nonpartisan congressional staff body whose conventional scores determine whether a reconciliation bill fits its instruction. Understanding how those scores are built is a subject of its own, treated in how CBO scores legislation, and the 2017 act is the case study that makes the mechanics concrete.

The headline estimate came in JCT publication JCX-67-17, released December 18, 2017, four days before signature. The Joint Committee projected that the conference agreement would reduce federal revenues by a net $1.456 trillion over fiscal years 2018 through 2027, including outlay effects, on a conventional basis. Conventional here means the estimate holds the size of the economy fixed and counts only the mechanical effect of the tax changes plus the behavioral responses of taxpayers, not any feedback from faster economic growth. The $1.456 trillion figure fit inside the $1.5 trillion reconciliation instruction with $44 billion to spare, which is why the conference rate landed at 21 percent rather than 20 and why the individual sunsets were drawn exactly where they were. The score was not a forecast of what would happen. It was the binding constraint the drafters had to satisfy, and they satisfied it by construction.

Other estimators produced larger numbers under different conventions. The Penn Wharton Budget Model, in its December 18, 2017 analysis of the conference agreement, estimated a static revenue loss of $2.209 trillion over 2018 through 2027 on a revenue-only basis, or $1.968 trillion including outlay effects, with dynamic feedback from economic growth lifting GDP by 0.6 to 1.1 percent by 2027 under its modeling assumptions. The differences between the Penn Wharton and JCT figures reflect different baselines, different behavioral assumptions and different treatments of outlays, not disagreement about the statute’s text. Readers comparing estimates should always check three things: the window, the baseline, and whether growth feedback is included. Estimates that differ on any of the three are not directly comparable, and much confused commentary about the act’s cost comes from comparing figures that were never meant to be compared.

What did the $1.456 trillion figure actually measure?

It measured the Joint Committee on Taxation’s conventional estimate of the conference agreement’s net budget effect over fiscal years 2018 through 2027, counting revenue changes and outlay effects while holding the economy’s size fixed. It excluded any growth dividend, which is why supporters called it understated and why it functioned as the binding reconciliation constraint.

The outlay side of the ledger deserves emphasis because tax debates usually ignore it. The $1.456 trillion net figure combined revenue losses with outlay changes, and the largest outlay change was the mandate zeroing: reducing the shared responsibility payment to zero was projected to lower federal spending on premium subsidies and Medicaid, because fewer people would carry qualifying coverage. Those outlay savings, scored in the hundreds of billions over the window, directly offset the revenue cost of the tax cuts in the official estimate. A reader who looks only at the revenue loss misses how the bill paid for part of itself within the score: not through growth, but through reduced health coverage spending. The mechanism was fully visible in the JCT tables. It was also, as the distributional section showed, the source of the finding that lower-income groups were net contributors to deficit reduction. The score and the distribution were two views of the same arithmetic.

Did the tax cuts pay for themselves?

No. The Joint Committee on Taxation projected a $1.456 trillion revenue loss over fiscal years 2018 to 2027, and post-enactment comparisons show fiscal year 2018 corporate receipts about $135 billion below the pre-enactment projection. Growth offset only a fraction of the static cost, leaving the self-financing claim unsupported by the measured record.

The post-enactment record on revenues is now long enough to check against the projections. A Brookings Institution analysis comparing actual fiscal year 2018 receipts with the Congressional Budget Office’s pre-enactment projections found corporate income tax revenue $135 billion below projection, a decline of roughly 40 percent, and individual income tax revenue $97 billion, or 5.4 percent, below projection. Total federal revenues came in at 16.4 percent of gross domestic product against 18.1 percent projected. The analysis included a telling control case: payroll tax receipts, which the statute did not touch, fell only 1.7 percent against projection, suggesting that the shortfall in the taxes the statute did touch was substantially attributable to the statute rather than to the economy. The revenue loss the estimators projected is, within the normal uncertainty of such comparisons, the revenue loss that occurred. The claim that the cuts would generate enough growth to pay for themselves, advanced during the debate by Treasury Secretary Mnuchin among others, is not consistent with this record. Growth feedback offset only a fraction of the static cost. That is a measured result, not a talking point, and it belongs in any honest account of the statute.

The scoring process itself deserves a final note, because the 2017 act was the most consequential test the process had ever faced. The Joint Committee’s conventional estimate is the number that legally matters: it determines whether a reconciliation bill fits its instruction, and no bill that fails the estimate can proceed. But conventional scoring deliberately excludes macroeconomic feedback, which meant the official $1.456 trillion figure did not credit the bill with any growth dividend at all. Supporters argued this understated the bill’s fiscal virtue; critics argued that crediting speculative growth would let Congress spend the same dollar twice. In the end the bill was designed to the conventional score, passed on the conventional score, and judged by history against the conventional score’s predictions. The dynamic debate changed no vote and altered no provision. It was, like the sunsets, part of the theater that the arithmetic required.

Who gained, and when the gains faded

Distributional estimates ask a different question from revenue estimates: not how much the government lost, but who received the benefit. The standard measure is the change in after-tax income by income group, and the standard sources are the Tax Policy Center, the Congressional Budget Office and the Joint Committee on Taxation, each using its own incidence assumptions about who ultimately bears the corporate tax. Those assumptions matter. The estimators generally assign most of the corporate tax to owners of capital: the Congressional Budget Office and the Joint Committee assume roughly 75 percent falls on capital and 25 percent on labor, the Tax Policy Center allocates 20 percent to workers and 80 percent to capital, and the Treasury has used roughly 18 percent to workers and 82 percent to capital. Because the statute’s largest dollar item was a corporate rate cut, the choice of incidence assumption drives the distributional result, and readers should know which assumption sits under any number they are shown.

The Tax Policy Center’s December 18, 2017 analysis of the conference agreement found that taxes fell on average for every income group in 2018, with average after-tax income rising 2.2 percent. But the gains were concentrated: as a share of income, the largest cuts went to households in the 95th through 99th percentiles. And the time pattern is the distributional mirror of the reconciliation asymmetry. By 2027, with the individual provisions expired and only the permanent business provisions remaining, the overall reduction in taxes was just 0.2 percent of after-tax income. The benefits faded because the provisions conferring them were written to fade. The distributional story and the procedural story are the same story told in different units. The 95th-to-99th-percentile finding deserves a gloss: those are households wealthy enough to benefit substantially from the corporate rate cut through their capital ownership and from the individual rate cuts and the pass-through deduction, but not so wealthy that the estate tax doubling dominated their benefit. The very top of the distribution gained more in dollars but less as a share of income from the income-tax provisions, while benefiting most from the estate provision. Distributional analysis always depends on the unit of measurement, and the profile reports both.

The Congressional Budget Office’s distributional analysis, released December 21, 2017, made the point more sharply by asking which groups were net contributors to deficit reduction, meaning groups whose combined tax and benefit changes reduced the deficit rather than increasing it. In 2019 the groups with income under $20,000, about 23 percent of taxpayers, were net contributors, mainly because the zeroing of the shared responsibility payment reduced health coverage subsidies flowing to lower-income households. In 2021, 2023 and 2025 the groups under $40,000, about 43 percent of taxpayers, were net contributors. By 2027 the groups under $75,000, about 76 percent of taxpayers, were net contributors to deficit reduction. The mechanism is worth stating plainly: as the individual cuts expired, lower- and middle-income households lost their tax relief, while the permanent corporate provisions continued to deliver benefits concentrated among owners of capital, and the mandate zeroing’s subsidy reductions persisted. The CBO figures are not a moral judgment. They are the arithmetic of sunsets.

None of this was hidden at the time. The distributional tables were published before the final votes, and the sunset schedule was in the bill text. But the public debate largely proceeded as though the 2018 distribution would persist, which is another cost of the reconciliation asymmetry: a temporary distribution was sold, in much commentary, as though it were permanent. The statute’s defenders could point to the 2018 numbers and its critics to the 2027 numbers, and both were citing official estimates of the same law. The honest statement is the time path, not either endpoint.

The Joint Committee on Taxation published its own distributional analysis in JCX-68-17, released the same day as its revenue estimate, showing the change in federal taxes by income group under the conference agreement. The JCT tables, like the Tax Policy Center’s, showed broad-based reductions in the early years fading over time as the individual provisions expired. When estimators who use different models and different incidence assumptions agree about the shape of the distribution over time, the shape is driven by the statute’s sunsets rather than by modeling choices.

A final distributional subtlety concerns the mandate zeroing, whose effects ran through outlays rather than taxes. The Congressional Budget Office’s finding that lower-income groups were net contributors to deficit reduction rested substantially on the loss of health coverage subsidies: zeroing the penalty reduced coverage, which reduced subsidy outlays, which counted as deficit reduction. This is a peculiar kind of fiscal contribution, one that made the budget look better while making the affected households worse off, and it illustrates how reconciliation arithmetic can produce distributional results that no one would defend as policy.

Investment, wages, and the measured record

The debate over the 2017 act made large claims about investment and wages, and the years since have produced a genuine empirical literature testing them. This section reports what that literature found, with its uncertainties intact. The fair summary, and the one the complication in the brief requires, is that the corporate provisions produced a measurable but modest investment response and wage effects well below the projections offered during the debate, alongside substantial revenue loss. That sentence contains three findings and each needs its own evidence.

Start with the investment response. An International Monetary Fund working paper published in 2019 (Working Paper 19/120) compared the post-enactment investment path with what prior tax-cut episodes would have predicted. The predicted boost to GDP was 1.3 percentage points against an actual 0.7, and for business investment 5.2 percentage points predicted against 3.5 actual. Only about 20 percent of the incremental corporate free cash flow went to capital expenditure and research; the rest went to share repurchases, dividends and balance-sheet adjustments. The paper cited two dampeners: elevated policy and trade uncertainty, and corporate market power, which let firms with pricing power return cash to shareholders rather than invest it.

Firm-level announcements told a similar story in miniature. Economists Michelle Hanlon, Jeffrey Hoopes and Joel Slemrod examined public-company behavior in the first quarter of 2018 and found that 4 percent of public firms announced worker-related benefits tied to the act, such as bonuses or wage increases, while 22 percent of S&P 500 firms mentioned increased investment on earnings calls. Share repurchases rose, but they were extremely concentrated, and only nine firms explicitly attributed a new repurchase plan to the statute. There was no significant increase in dividends. The much-publicized announcements of worker bonuses, real as they were for the workers who received them, came from a small minority of firms. The concentration of the buybacks matters for interpreting the aggregate data: economy-wide repurchase figures rose substantially after enactment, but the firm-level attribution shows that most of the increase came from firms acting for reasons they did not connect to the statute, while the firms that did connect their actions to the statute were disproportionately spending their transition-tax windfall. Aggregate time series can mislead when the composition is this skewed, which is why the paper’s firm-level design added value beyond the national accounts.

A University of North Carolina study by Bennett, Thakor and Wang added an important causal distinction in August 2019. The post-enactment surge in share repurchases, they found, was driven by the repatriation tax cut under section 965 rather than by the income tax cut: firms used the newly accessible offshore cash to buy back shares. Investment, by contrast, increased primarily in response to the income tax cut, the rate reduction itself, and especially among capital-constrained, low-debt, profitable firms, the firms for which the cost of capital actually binds. The two headline uses of the windfall thus had different causes: buybacks came from the transition tax, investment from the rate cut. Lumping them together, as much commentary did, obscures the mechanism.

Why did wage gains fall short of the projections?

Corporate tax cuts flow mostly to owners of capital, and the official estimators assumed roughly three-quarters of the burden sits there, so only a modest share reaches paychecks. Matched employer-employee records show earnings gains concentrated among executives and the top tenth within firms, with no change for the bottom 90 percent.

The wage evidence is the most studied and the most sobering part of the record. A paper by Kennedy, Dobridge, Landefeld and Mortenson, published in the American Economic Review and using matched employer-employee tax records to compare C corporations, which received the rate cut, with S corporations, which did not, found that the corporate cut raised sales, profits, investment and employment. But the earnings gains were concentrated in executive compensation and the top 10 percent of the within-firm earnings distribution; workers in the bottom 90 percent saw no change in earnings. The authors’ estimates imply that each dollar of forgone corporate tax revenue generated about ten cents of additional output, with 78 percent of the gains accruing to the top tenth of the income distribution in the working paper’s accounting, and the published version emphasizing that roughly 87 percent of short-term private income gains went to the top tenth. The mechanism is the incidence assumption made flesh: if three-quarters of the corporate tax falls on capital, then cutting the corporate tax mostly enriches capital owners, and the wage channel is narrow by construction.

Set against this record the claims made during the debate, stated here in their strongest form, because fairness requires it. The Council of Economic Advisers under Chairman Kevin Hassett published a paper in October 2017 arguing that cutting the statutory corporate rate from 35 percent to 20 percent would raise average household income by “very conservatively” $4,000 a year, possibly up to $9,000. The mechanism was capital mobility: American firms shifted capital and profits abroad to chase lower rates, and a competitive domestic rate would bring capital home, raising domestic investment, labor demand and wages. Note the input: the paper modeled a 20 percent rate, while the enacted rate was 21 percent, a small difference that its critics noted and its authors would consider immaterial to the qualitative claim. Separately, an analysis by Boston University economist Laurence Kotlikoff, circulated by the Speaker’s office in October 2017, projected GDP gains of 3 to 5 percent and real wage gains of 4 to 7 percent, roughly $3,500 per working household annually. Treasury Secretary Mnuchin argued the cuts would generate enough growth to pay for themselves. The Tax Foundation’s dynamic model projected larger long-run GDP and wage gains under its own growth assumptions.

Each of these claims met pointed criticism at the time, and the criticism belongs in the record alongside the claims. An Initiative on Global Markets survey of leading economists found 37 of 38 agreeing that the act would rapidly increase the national debt, with only one, Stanford’s Darrell Duffie, agreeing that GDP would be substantially higher a decade out. Lawrence Summers called the $4,000 figure absurd on its face and said it had no peer-reviewed support. Harvard economist Mihir Desai said the Council had misinterpreted his own incidence research in building its case. An Economic Policy Institute paper argued there was no consistent relationship between post-tax profit rates and productivity-enhancing investment. And the measured outcomes summarized above, the IMF’s below-prediction investment path, the 4 percent of firms announcing worker benefits, the concentration of earnings gains in the top tenth, sit well below the debate’s projections. The uncertainty in this literature is real: identification is hard, the post-enactment period included trade disputes and then a pandemic that complicate any before-and-after comparison, and longer-run effects may differ from short-run ones. But the direction of the findings is consistent across independent research designs, and an honest profile reports the consistency along with the caveats. The growth argument was made in good faith and in its strongest form above; the measured results did not confirm it.

The incidence question that underlies the wage debate deserves one more paragraph, because it is where the economic theory and the political argument most directly collided. The Council of Economic Advisers’ case for large wage gains rested on the claim that the corporate tax is substantially borne by workers, through lower investment, lower productivity and lower wages. The official estimators’ assumption, that roughly three-quarters of the tax falls on owners of capital, implies the opposite: cutting the tax mostly benefits shareholders. Both claims cannot be right, and the difference between them is the difference between $4,000 per household and a fraction of that. The published incidence literature, surveyed before and after the act, generally supports the estimators’ view for a large economy with imperfect capital mobility, while acknowledging that openness to capital flows pushes some of the burden onto labor. Desai’s objection to the Council’s paper was precisely that it had taken his research on openness and read it as supporting a larger labor share than the research would bear. The wage results that followed, concentrated at the top of the within-firm distribution, are consistent with the estimators’ incidence assumption and difficult to square with the Council’s. Theory predicted modest wage effects. Measurement found them. The debate’s larger numbers remain, on the evidence, projections without a confirming record.

How 2017 differed from 1986

The 2017 act invites comparison with the Tax Reform Act of 1986, the last comprehensive rewrite of the code, and the comparison clarifies what was genuinely new. The series treats the structural comparison at length in Tax Reform 1986 vs. TCJA 2017, so this section states only the points the comparison needs. The 1986 reform was revenue-neutral by design, bipartisan in passage, and enacted through regular order without reconciliation; it paid for lower rates by broadening the base, eliminating shelters and preferences, and it was permanent throughout. The 2017 act was none of those things: it was scored as a $1.456 trillion revenue loss, passed without a minority-party vote in either chamber’s final tally, moved through reconciliation, broadened the base only partially, and split its provisions between permanent and expiring. The 1986 act’s durability came from consensus. The 2017 act’s split durability came from the Byrd rule. That is the difference between a reform built to last and a reform built to fit.

The contrast extends to the international titles, where the two reforms moved in opposite directions from different starting points. In 1986 the United States strengthened its worldwide system, adding the passive foreign investment company rules and tightening the foreign tax credit, because the problem of the era was seen as insufficient taxation of offshore income. In 2017 the United States abandoned the worldwide system for a territorial one, because the problem of the era was seen as the lockout effect and the competitive disadvantage of the high rate. The two reforms agreed that the prior international system was broken and disagreed about the direction of the fix, which is a reminder that tax policy’s pendulum swings with the perceived problem rather than toward any permanent wisdom. The 2017 act’s territorial turn, policed by GILTI, FDII and BEAT, was the more technically elaborate of the two, and its elaborateness reflects how much harder it has become to tax mobile income in an integrated world economy.

There is also a procedural moral in the comparison. The 1986 act took more than two years from proposal to signature, moved through regular order with extensive hearings, and passed the Senate 97 to 3. The 2017 act took fifty days from introduction to signature, moved through reconciliation with limited debate, and passed the Senate 51 to 48. The first process produced a statute that endured for decades with broad legitimacy. The second produced a statute whose most prominent provisions carried expiration dates and whose name was struck before enactment. Speed has costs, and the 2017 act itemizes them. Whether the trade was worth it is a judgment for the reader. That the trade was made, and that its terms are written into the statute’s structure, is the historical fact.

Verdict: procedure wrote the text

Return to the three abilities this profile promised. A reader who has worked through it can explain why the individual provisions expire while the corporate rate cut does not: reconciliation forbids deficit increases beyond the budget window, the drafters sunset the individual provisions to satisfy that rule, and the corporate cut fit the window without a sunset. The reader can state the procedural rule that forced the asymmetry: the Byrd rule, section 313 of the Congressional Budget Act, which polices reconciliation bills for extraneous provisions and out-year deficit increases, and whose point-of-order process deleted the bill’s own name on December 19, 2017. And the reader knows the fact that almost no coverage reports: the short title “Tax Cuts and Jobs Act” was struck from the bill before enactment, so the law enrolled as Public Law 115-97 carries a descriptive reconciliation title and is not formally named the Tax Cuts and Jobs Act at all.

The reconciliation asymmetry is the profile’s namable claim and its verdict at once. The durability of every provision in the 2017 act was determined by a Senate procedural rule rather than by any judgment about which policies deserved to last. The individual rate cuts, the centerpiece of the public case for the bill, expire. The corporate alternative minimum tax repeal, which few voters could have described, is permanent. The pass-through deduction expires while the corporate rate cut endures, so two businesses with identical income face different long-run tax treatment based solely on legal form. None of these outcomes expresses a ranking of policies. All of them express the scoring window.

The series thesis appears here at its sharpest and most recent point: procedure determining text. Earlier statutes in the series were shaped by procedure at the margins. The 2017 act was shaped by procedure in its architecture. The sunsets are procedure. The struck title is procedure. The 21 percent rate, chosen over 20 to fit the instruction, is procedure. The permanent-or-expiring table above is, in effect, a map of the Byrd rule’s shadow across the Internal Revenue Code. The 2017 act is the limiting case, the statute in which the procedure is the text, and it earns its place in the series as the clearest demonstration of the thesis.

Before the final cautions, a checklist of the four near-misses this profile was built to prevent, stated once more in compact form for the reader who will cite the statute. First, do not assume the whole act is permanent. Roughly half of its headline provisions expire after 2025, and the permanent half is mostly the business and international titles. Second, do not read the sunsets as policy. The expiration dates were a compliance device for the Byrd rule’s out-year test, not an expression of doubt about the individual provisions or a ranking of their importance. Third, do not cite the popular name as the formal title. The short title “Tax Cuts and Jobs Act” was struck on a point of order on December 19, 2017, and the enrolled law bears the descriptive reconciliation title; the name everyone uses is colloquial, not enacted. Fourth, do not repeat the drafting-stage numbers as law. The enacted corporate rate is 21 percent, not the 20 percent of the pre-conference bills. The enacted mortgage ceiling is $750,000, not the $500,000 of the House bill. The personal exemption was suspended to zero, not repealed. The mandate was zeroed, not repealed. Each of these errors has a specific origin in the legislative history, and each is corrected by reading the enrolled text rather than the debate about it.

Two final cautions for the reader who will use this profile as a reference. First, the expiration dates and rate schedules describe the statute as enacted in December 2017, as noted with the permanent-or-expiring table above. Nothing in this article addresses any proposal to extend, modify or repeal any provision; the profile is a history of the enactment, not a brief for any future action. Second, the popular name remains the right search term and the wrong citation. Call it the Tax Cuts and Jobs Act in conversation and in headlines, as nearly everyone does, but never cite it as the law’s enacted short title, never assume its provisions share a single lifespan, and never read the sunsets as policy. The law is Public Law 115-97. Its name was struck. Its structure was scored. And its most important author may have been a Senate rule.

Readers who want to work this material into their own study notes can build an outline from the table and the provision groups using VaultBook’s legislation study notebook, which is designed for exactly this kind of statute profile.

Frequently Asked Questions

Q: What did the Tax Cuts and Jobs Act change?

The act rewrote the individual, business, and international tax rules in a single reconciliation bill, Public Law 115-97, signed December 22, 2017. For individuals, it lowered rates to a top bracket of 37 percent, nearly doubled the standard deduction to 12,000 dollars single and 24,000 dollars joint, suspended the personal exemption, capped the state and local tax deduction at 10,000 dollars, limited mortgage interest to 750,000 dollars of new debt, doubled the estate tax exemption to about 11.18 million dollars, and reduced the health coverage penalty to zero. For businesses, it cut the corporate rate permanently from 35 to 21 percent, repealed the corporate alternative minimum tax, expanded bonus depreciation with a phase-down, limited interest deductions to 30 percent of adjusted taxable income, and created a 20 percent deduction for pass-through income. Internationally, it moved toward a territorial system with a participation exemption, a one-time transition tax, and the GILTI, FDII, and BEAT regimes. Nearly all individual provisions expire after 2025; the corporate and international provisions are permanent.

Q: Why do Tax Cuts and Jobs Act individual provisions expire?

The individual provisions expire after December 31, 2025, because of the Byrd rule, the Senate procedural rule governing budget reconciliation bills. The Byrd rule allows any senator to strike provisions that increase federal deficits beyond the budget window covered by the budget resolution, which for this bill was fiscal years 2018 through 2027. Permanent individual tax cuts would have increased deficits beyond that window, inviting a fatal point of order that needed sixty votes to waive. The bill’s writers therefore gave the individual provisions expiration dates, booking only eight years of their cost instead of ten. The corporate rate cut needed no sunset because its ten-year cost fit inside the 1.5 trillion dollar reconciliation allowance. The asymmetry between expiring individual provisions and permanent corporate ones reflects Senate procedure, not a judgment that corporate relief deserved permanence while family relief deserved only a trial run.

Q: What is the SALT cap in the Tax Cuts and Jobs Act?

The SALT cap limits the federal itemized deduction for state and local taxes to 10,000 dollars per household, 5,000 dollars for married taxpayers filing separately, for tax years 2018 through 2025. It covers the aggregate of state and local income taxes (or sales taxes, at the taxpayer’s election) plus property taxes. Taxes incurred in carrying on a trade or business remain deductible outside the cap. Before the cap, itemizers could deduct the full amount of state and local taxes paid, which meant the federal government effectively subsidized part of high state and local tax burdens. The cap ended that open-ended subsidy above 10,000 dollars. Its burden falls hardest on higher-income itemizers in high-tax states, where combined property and income tax bills routinely exceed the ceiling. It sunsets with the other individual provisions after 2025. Several states sued over the cap and lost, tried charitable-contribution workarounds that regulators disallowed, and enacted entity-level pass-through taxes that partially blunted it.

Q: What is section 199A in the Tax Cuts and Jobs Act?

Section 199A is the 20 percent deduction for qualified business income from pass-through entities, meaning partnerships, S corporations, and sole proprietorships. It lets eligible owners deduct up to 20 percent of domestic business income below the line, without itemizing, capped at 20 percent of taxable income computed before the deduction and excluding net capital gains. Above statutory income thresholds, two limits phase in: a wage and capital limit tied to W-2 wages paid and property investment, and an exclusion for specified service trades or businesses such as health, law, accounting, consulting, and financial services, with engineers and architects carved out. Congress created it to narrow the rate gap between the 21 percent corporate rate and the 37 percent top individual rate, which otherwise would have pushed pass-throughs to convert to C corporations. It is scored as an individual provision and sunsets after 2025, even though its economic rationale does not expire.

Q: Is the Tax Cuts and Jobs Act its official name?

No. The short title “Tax Cuts and Jobs Act” was struck from the bill on a Byrd-rule point of order on December 19, 2017, before final passage, so the enacted law is not formally named the Tax Cuts and Jobs Act at all. The Senate parliamentarian had advised that the short title was extraneous under the Byrd rule because a name has no budgetary effect, Senator Bernie Sanders raised the point of order, Senator Mike Enzi’s motion to waive the rule failed 51 to 48 with sixty votes required, and the title fell along with two substantive provisions. The enrolled law’s official title is the descriptive reconciliation title: “An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018.” Everyone from the Internal Revenue Service to the press continued calling it the Tax Cuts and Jobs Act, but citing the popular name as the enacted short title misstates the statute’s identity.

Q: What did the Tax Cuts and Jobs Act do to the corporate rate?

The act replaced the graduated corporate rate structure topping out at 35 percent with a flat 21 percent rate, effective for tax years beginning after December 31, 2017. The cut is permanent: the statute contains no sunset for the 21 percent rate. Both chambers’ pre-conference bills had set the rate at 20 percent, and the conference agreement moved it to 21 percent when the final budget arithmetic needed room under the 1.5 trillion dollar reconciliation cap, so summaries citing 20 percent describe the drafts rather than the law. The act also repealed the corporate alternative minimum tax outright, with transition rules letting corporations use accrued AMT credits. The permanent corporate rate contrasts with the expiring individual rate cuts, and the contrast is procedural: the corporate cut’s ten-year cost fit inside the budget window without a sunset, while permanent individual cuts would have breached the Byrd rule’s bar on deficit increases beyond the window.

Q: How did the Tax Cuts and Jobs Act pass the Senate?

The act passed the Senate through budget reconciliation, which allowed passage with a simple majority instead of the sixty votes needed to overcome a filibuster. The Senate first passed its version in the early morning of December 2, 2017, by 51 to 49, with Senator Bob Corker of Tennessee the only Republican voting no, on deficit grounds. No vice-presidential tie-break was needed; the only Pence tie-break in the process, 51 to 50, decided the Cruz amendment on 529 accounts during the vote-a-rama, not the bill itself. After the conference report, the Senate parliamentarian advised that three provisions violated the Byrd rule, Senator Bernie Sanders raised points of order, the waiver motion failed 51 to 48, and the provisions were stricken. The Senate then passed the amended bill on December 20 by 51 to 48, Record Vote 323, with Senator John McCain absent. The House concurred the same day, 224 to 201.

Q: Did the Tax Cuts and Jobs Act raise wages?

The measured record does not support broad wage gains. The Council of Economic Advisers had claimed in October 2017 that cutting the corporate rate would raise average household income by “very conservatively” 4,000 dollars per year, but that projection rested on contested assumptions about how much of the corporate tax falls on workers. The most granular test came from a study published in the American Economic Review that matched employer-employee tax records and compared C corporations, which got the rate cut, against S corporations, which did not. It found the cut raised sales, profits, investment, and employment, but earnings gains were concentrated in executive pay and the top tenth of each firm’s pay distribution, while workers in the bottom 90 percent saw no earnings change. A separate study found only about 4 percent of public firms announced worker benefits tied to the act. Investment rose but below historical predictions. The honest summary is measurable but modest investment effects and wage gains well below the debate’s projections.

Q: What is the GILTI regime in the Tax Cuts and Jobs Act?

GILTI, the tax on global intangible low-taxed income under section 951A, is the act’s backstop against profit shifting under the new territorial-leaning system. Despite the name, it reaches beyond intangibles: it imposes current-year United States tax on a domestic corporation’s net controlled foreign corporation income exceeding a 10 percent return on the tangible depreciable assets used in the foreign business. The 10 percent allowance represents a deemed routine return on physical investment, and everything above it is treated as mobile, low-taxed income. The effective rate is 10.5 percent for corporate taxpayers, achieved through a 50 percent deduction under section 250. Because the routine-return allowance is measured against tangible assets, asset-light businesses with large foreign earnings face GILTI on most of their foreign income, while asset-heavy manufacturers shelter more. The provision is permanent, though its deduction is scheduled to narrow after 2025, raising the effective rate.

Q: What is the FDII deduction in the Tax Cuts and Jobs Act?

FDII, the deduction for foreign-derived intangible income under section 250, is the carrot paired with GILTI’s stick. It provides a preferential effective rate of 13.125 percent, through a 37.5 percent deduction against the 21 percent corporate rate, on the portion of a domestic corporation’s deemed intangible income attributable to foreign sales and services. Where GILTI punishes booking mobile income in low-tax foreign jurisdictions, FDII rewards booking it at home: a corporation that holds its intellectual property in the United States and sells abroad pays a rate well below 21 percent on the resulting income. The two provisions were calibrated as a pair, with GILTI’s 10.5 percent and FDII’s 13.125 percent creating a mild preference for domestic over foreign booking of mobile income. Like GILTI, the FDII deduction is scheduled to narrow after 2025, raising its effective rate. The provision is permanent in structure.

Q: What is the BEAT in the Tax Cuts and Jobs Act?

BEAT, the base erosion and anti-abuse tax under section 59A, is a minimum tax aimed at deductible payments from domestic corporations to foreign affiliates. It applies to large taxpayers with average annual gross receipts of at least 500 million dollars over three years, whose base erosion percentage, roughly the share of deductions attributable to payments to foreign related parties, reaches 3 percent, or 2 percent for banks and registered securities dealers. The BEAT rate was 5 percent for 2018, rose to 10 percent for 2019 through 2025, and was scheduled to reach 12.5 percent after 2025, with banks and dealers facing rates one point higher at each step. It operates by recomputing taxable income with base erosion payments added back and applying the BEAT rate to that broader base, collecting any excess over regular tax liability. It was the bluntest instrument in the international title, sweeping in ordinary cross-border payments alongside abusive ones.

Q: What was the section 965 transition tax in the Tax Cuts and Jobs Act?

Section 965 imposed a one-time tax on the accumulated post-1986 foreign earnings of specified foreign corporations, deemed repatriated in the last tax year beginning before January 1, 2018. Earnings held in cash or cash equivalents were taxed at 15.5 percent; the remainder at 8 percent. The measurement date was the greater of the earnings and profits measured on November 2, 2017 or December 31, 2017, and the resulting liability could be paid in eight back-loaded installments. The provision cleared the vast stockpile of unrepatriated foreign earnings, estimated above 2.5 trillion dollars, that had accumulated under the old worldwide system with deferral, so the new participation exemption could start from a clean slate. Research on the aftermath found the freed cash flowed substantially to share repurchases rather than new domestic investment: one study attributed the post-act buyback surge to the repatriation cut rather than the income-tax cut. The tax applied once and then retired.

Q: How did bonus depreciation phase down under the Tax Cuts and Jobs Act?

The act allowed 100 percent bonus depreciation, immediate expensing, for qualified property acquired and placed in service after September 27, 2017 and before January 1, 2023. The percentage then steps down on a statutory schedule: 80 percent for property placed in service in 2023, 60 percent in 2024, 40 percent in 2025, 20 percent in 2026, and zero in 2027, with long-production-period property on the same schedule delayed by one year. The September 2017 start date reached back before enactment to the release of the House leadership’s tax framework, rewarding investment made in anticipation of the law. Unlike the individual sunsets, which were Byrd-rule compliance devices, the phase-down was a designed cost-control schedule: it let the writers claim the full incentive effect up front while booking a declining cost over the budget window. Economists debate whether scheduled phase-downs increase total investment or merely shift its timing, since firms expecting the step-down accelerate purchases into the 100 percent years.

Q: What did section 163(j) limit in the Tax Cuts and Jobs Act?

Section 163(j) limits a business’s deduction for business interest to the sum of its business interest income, 30 percent of its adjusted taxable income, and its floor plan financing interest, effective for tax years beginning after December 31, 2017. Small businesses are exempt if average annual gross receipts are 25 million dollars or less over the prior three years, indexed for inflation. Disallowed interest carries forward indefinitely rather than being lost. The definition of adjusted taxable income tightened over time: before 2022 it was computed on a basis comparable to earnings before interest, taxes, depreciation, and amortization, while starting in 2022 it shifted toward earnings before interest and taxes, reducing the 30 percent allowance and increasing the limit’s bite. The provision was designed as permanent base protection paired with the lower corporate rate, ensuring heavily indebted firms could not deduct their way to near-zero liability at 21 percent. It has no sunset.

Q: What happened to the estate tax exemption in the Tax Cuts and Jobs Act?

The act doubled the federal estate tax basic exclusion amount to 11,180,000 dollars per decedent for 2018, up from 5,490,000 dollars in 2017, with inflation indexing in later years. In practical terms, the change removed all but the very largest estates from the federal estate tax for the provision’s duration. Like the other individual provisions, the doubling sunsets after December 31, 2025, a Byrd-rule artifact rather than a policy judgment: a permanent doubling would have increased deficits beyond the budget window. The scheduled reversion created a use-it-or-lose-it planning environment, as advisers urged clients with estates between the old and new thresholds to make large gifts before the deadline. The central legal question was whether gifts made under the doubled exemption would be clawed back if the donor died after it fell, and Treasury regulations answered no, which itself encouraged a wave of large gifts in the window’s later years. The sunset generated its own body of planning law.

Q: Who raised the Byrd-rule points of order against the tax bill?

Senator Bernie Sanders of Vermont, the ranking member of the Senate Budget Committee, raised the points of order on December 19, 2017, in a joint statement with Senator Ron Wyden of Oregon, the ranking member of the Finance Committee. The Senate parliamentarian had advised that three provisions were extraneous under the Byrd rule: the short title “Tax Cuts and Jobs Act,” the Cruz provision expanding section 529 accounts to homeschooling expenses, and part of the endowment excise tax’s small-institution exception. Senator Mike Enzi of Wyoming moved to waive the Byrd rule, which required sixty votes; the motion failed 51 to 48, and the three provisions were stricken. The attribution matters because a common error credits the parliamentarian alone with striking the title. The parliamentarian advised, a ranking member chose to raise the points, the majority tested the waiver, and the Senate’s arithmetic did the rest. Procedure determined the text, but people operated the procedure.

Q: How did the Tax Cuts and Jobs Act change the individual mandate?

The act did not repeal the Affordable Care Act’s individual mandate; it reduced the shared responsibility payment, the penalty for lacking health coverage, to zero for tax year 2019 and all subsequent years, in section 11081. The mandate’s statutory language survived intact, which is why careful descriptions say the penalty was zeroed rather than the mandate repealed. The distinction carried constitutional consequences litigated afterward over whether the unenforceable mandate could be severed from the rest of the health law. Within the tax story, the zeroing mattered for two reasons. First, it was scored as reducing federal outlays, because the Congressional Budget Office projected fewer people would enroll in subsidized coverage without the penalty, and that outlay reduction helped the bill fit the reconciliation cap. Second, unlike the other individual provisions, the zeroing was permanent: reducing a penalty lowers outlays rather than raising deficits, so the Byrd rule’s deficit test gave it a pass.

Q: How did the Tax Cuts and Jobs Act change the distribution of taxes?

The Tax Policy Center’s analysis of the conference agreement found taxes falling on average for all income groups in 2018, with average after-tax income rising 2.2 percent, but the largest cuts as a share of income went to the 95th through 99th percentiles. By 2027, after the individual provisions expired, the overall tax reduction was only 0.2 percent of after-tax income, and the remaining benefit flowed mostly to capital owners through the permanent corporate provisions. The Congressional Budget Office’s distributional analysis added that the mandate zeroing’s outlay effects made lower-income groups net contributors to deficit reduction: in 2019, groups under 20,000 dollars of income were net contributors, and by 2027, groups under 75,000 dollars, about 76 percent of taxpayers, were. The pattern reflects the reconciliation asymmetry directly: the expiring provisions delivered most of the middle-class benefit, while the permanent provisions delivered most of the lasting benefit to capital. Averages within groups mask wide individual variation.

Q: What was the Byrd bath for the 2017 tax bill?

The Byrd bath is the pre-floor review in which the Senate parliamentarian examines each provision of a reconciliation bill against the six tests of the Byrd rule and advises which ones are extraneous. The advice is technically advisory but practically decisive, because any senator can raise a point of order against a flagged provision and overcoming it requires sixty votes to waive the rule. For the 2017 tax bill, the parliamentarian advised that three provisions were extraneous: the short title “Tax Cuts and Jobs Act,” which had no budgetary effect; the Cruz provision expanding section 529 accounts to homeschooling; and the endowment excise tax’s exception for institutions with fewer than 500 tuition-paying students. Senator Bernie Sanders raised the points of order on December 19, 2017, Senator Mike Enzi’s waiver motion failed 51 to 48, and all three provisions were stricken, forcing the House to revote. The bath is where reconciliation bills are edited in substance.

Q: How did states respond to the SALT cap?

States responded to the 10,000 dollar SALT cap with lawsuits, workarounds, and new taxes of their own. New York, New Jersey, Connecticut, and Maryland sued the federal government, arguing the cap unconstitutionally intruded on state taxing authority; the Southern District of New York dismissed the suit and the Second Circuit affirmed, ending the challenge. New York, New Jersey, and Connecticut created charitable-contribution programs designed to let taxpayers recharacterize state tax payments as deductible charitable gifts, but Treasury and IRS regulations disallowed the maneuver. The most durable response was the pass-through entity tax: states imposed entity-level taxes on partnerships and S corporations, deductible at the entity level outside the SALT cap, with offsetting credits to owners, and a large majority of states eventually enacted some version. The episode showed the cap working as designed while generating the avoidance industry its critics predicted, with the workarounds narrowing but not removing its bite.