Tax reform is the rare legislative prize that Congress has captured twice in living memory under rules that could hardly be more different. The first capture, Public Law 99-514, the Tax Reform Act of 1986, lowered the top statutory individual rate from 50 percent to 28 percent and the corporate rate from 46 percent to 34 percent while holding federal revenue roughly constant. The 1986 act cleared the Senate 97 to 3 on its pre-conference passage and the final conference report 292 to 136 in the House and 74 to 23 in the Senate, with majorities of both parties in favor; the 2017 act, Public Law 115-97, passed 51 to 48 in the Senate and 224 to 201 in the House on party-line votes. The second capture cut the corporate rate from 35 percent to 21 percent permanently, lowered individual rates only through 2025, and was estimated by the Joint Committee on Taxation to increase the deficit by about $1.46 trillion over ten years on a conventional score. This article compares the two statutes across four axes, fiscal design, distributional design, procedure, and durability, and then delivers a defended verdict on which model a future reform should follow.

Students, journalists, and staffers routinely reach for this comparison and almost always receive it in slogan form: the 1986 act was bipartisan and the 2017 act was not, the 1986 act was paid for and the 2017 act was not, the 1986 act lasted and the 2017 act is expiring. Each slogan contains a grain of accuracy and obscures the mechanism underneath. The more useful question is structural. Why did one design produce a supermajority while the other produced a party-line vote? Why did one statute durably change the tax base while the other durably changed a rate? And which pattern should a future Congress copy if it wants a reform that survives contact with the next election? Answering those questions requires treating the two laws as products of four distinct choices, and it requires resisting the temptation to compare headline rate cuts without comparing the base changes, the procedures, and the durability strategies that sat underneath them.

The four axes do the organizing work. Fiscal design asks how each statute handled the budget constraint: revenue neutral by construction, or enacted under an instruction that permitted a large revenue loss. Distributional design asks whether each statute held the distribution of the burden constant as a binding constraint, or whether it left distribution free to move. Procedure asks how each statute moved through Congress: regular order through both tax-writing committees and a conference, or reconciliation with a 51-vote threshold and no realistic prospect of minority support. Durability asks what survived: whose base broadening endured, whose rates endured, and what the asymmetry reveals about how tax law actually persists. Readers who can state the four differences, explain the two coalitions, say which statute changed the base and which changed the rates, and defend a verdict on the better model have passed the article’s one test. Everything that follows exists to make that possible.

A note on method and fairness belongs up front. This comparison restricts its judgment to structural durability and coalition-building, and it takes no position on which distribution of the tax burden is preferable. Every distributional estimate in the article is attributed to a named source with its window, and readers who prefer a different distribution of the burden should read the verdict as a claim about legislative mechanics instead of a claim about fairness. The numbers come from the Joint Committee on Taxation, the Congressional Budget Office, and the Tax Policy Center, and the procedural facts come from the Congressional Record and the parliamentarian’s rulings. Where estimates differ across sources, the article says so instead of smoothing the disagreement away.

The article’s structure mirrors the one test it sets. The opening sections establish what each statute contained and how each reached the president’s desk, so the comparison rests on enacted text instead of on reputation. The four-axis sections then work through fiscal design, distributional design, procedure, and durability in turn, each ending in the practical consequence the table summarizes. Two interpretive sections follow: one on the two theories of reform that the statutes embody, base first versus rate first, and one on why the vote counts looked so different, which synthesizes the axes into the coalition explanation. A section on the afterlife of each statute tests the durability claims against what later Congresses actually did. The complication section confronts the claim that the 2017 act was merely the 1986 act with different numbers, the objections section steelmans the best arguments against the verdict, and the verdict section names the deciding factor. A study section closes with a worked self-test. Readers in a hurry can read the table, the verdict, and the FAQs; readers who want the mechanism should read straight through.

One more preliminary point will prevent the most common misreading. This article compares legislative designs, not presidential administrations or parties. The 1986 act was signed by a Republican president and passed by a Democratic House and a Republican Senate; the 2017 act was signed by a Republican president and passed by Republican majorities in both chambers. Nothing in the comparison turns on which party held the pen. The same four axes would apply if the parties were reversed, and the verdict would come out the same way, because the verdict is about constraints and procedures instead of about partisanship. A Democratic majority that passed a deficit-financed, party-line tax bill through reconciliation would have paid the same price in durability and coalition breadth, and a Republican majority that accepted revenue and distributional neutrality under regular order would have bought the same structural durability. The neutrality price does not check party registration.

Why Tax Reform Passed Twice, and Differently

Before the axes, the statutes themselves need clear identities, because the comparison collapses the moment a reader confuses what each law actually contained. Public Law 99-514, signed on October 22, 1986, was the Tax Reform Act of 1986, and its enacted shape is recorded in detail in the complete guide to the Tax Reform Act of 1986. The statute collapsed the individual rate schedule into two statutory rates, 15 and 28 percent, with a 5 percent surcharge that produced an effective 33 percent top marginal rate, and it cut the corporate rate from 46 percent to 34 percent. It paid for those rate cuts by broadening the base: it repealed the investment tax credit, repealed the 60 percent exclusion for long-term capital gains so that gains were taxed as ordinary income, imposed strict limits on passive activity losses and at-risk deductions, repealed the deduction for consumer interest, expanded the alternative minimum tax, and curtailed an array of industry-specific preferences. At the same time it removed millions of low-income filers from the rolls by nearly doubling the standard deduction, enlarging the personal exemption, and expanding the earned income tax credit. The Joint Committee on Taxation estimated the individual provisions would lose about $122 billion and the corporate provisions would gain about $120 billion over five years, for a net of roughly negative $2 billion, which is essentially revenue neutral, and that construction meant every dollar of rate reduction had to be matched by a dollar of base broadening or revenue raised elsewhere in the bill. The Senate passed its version 97 to 3 on June 24, 1986, before the conference; the House had passed its version by voice vote on December 17, 1985. The final conference report passed the House 292 to 136 on September 25, 1986, and the Senate 74 to 23 on September 27, 1986, with majorities of both parties in favor, and a Democratic House, a Republican Senate, and a Republican president each owned a share of the result.

Public Law 115-97, signed on December 22, 2017, is the statute most readers know as the Tax Cuts and Jobs Act, and its provisions are walked through provision by provision in the guide to the Tax Cuts and Jobs Act. The comparison needs its shape stated precisely, because the statute mixed permanent and temporary provisions in a way the 1986 act never did. On the corporate side, the headline change was permanent: the top corporate rate fell from 35 percent to 21 percent with no expiration date. Around that rate cut sat a redesigned international system, with a one-time transition tax on deemed repatriated foreign earnings under section 965, a minimum tax on global intangible low-taxed income known as GILTI, a deduction for foreign-derived intangible income known as FDII, and a base erosion and anti-abuse tax known as BEAT. On the individual side, nearly everything was temporary: the top individual rate fell from 39.6 percent to 37 percent through 2025, the standard deduction roughly doubled through 2025, the personal exemption was suspended to zero through 2025, the deduction for state and local taxes was capped at $10,000 through 2025, and the section 199A deduction for qualified pass-through business income ran through 2025. The JCT’s revenue table for the ten-year window put the corporate rate cut at a loss of about $1,348.5 billion, the individual rate cuts at a loss of about $1,214.2 billion, the $10,000 SALT cap at a gain of about $670.6 billion, the section 199A deduction at a loss of about $414.5 billion, the section 965 transition tax at a gain of about $338.8 billion, GILTI at a gain of about $112.4 billion, the FDII deduction at a loss of about $63.7 billion, and BEAT at a gain of about $149.6 billion, all per CRS R48485. Not every individual-side item expired: the shift to chained CPI for indexing brackets and thresholds was permanent, 100 percent expensing for equipment began phasing down in 2022 and phases out by 2026, and the GILTI, FDII, and BEAT provisions step up after 2025. The Joint Committee on Taxation estimated in JCX-67-17 that the conference agreement would increase the deficit by about $1.46 trillion over fiscal years 2018 through 2027 on a conventional score, excluding macroeconomic feedback. The Senate passed the bill 51 to 48 on December 20, 2017, the House passed it 224 to 201 the same day, and no member of the minority party voted for it in either chamber. One procedural footnote deserves attention because it symbolizes the whole contrast: On December 19, 2017, a Byrd-rule point of order raised by Senator Sanders struck the bill’s short title, and a motion by Senator Enzi to waive the rule and preserve the title failed 51 to 48, so the statute that history calls the Tax Cuts and Jobs Act is formally just Public Law 115-97, an act without an official name.

The roads the two statutes traveled to the president’s desk could hardly be more different, and the contrast is worth narrating because it previews all four axes. The 1986 act grew out of a multi-year bipartisan conversation. President Reagan made tax reform a second-term priority and the Treasury Department produced a reform blueprint in 1984; in Congress, bipartisan bills associated with Senator Bill Bradley and Representative Dick Gephardt on the Democratic side and Representative Jack Kemp and Senator Bob Kasten on the Republican side kept competing versions of base-broadening, rate-cutting reform in front of the tax-writing committees. The House Ways and Means Committee under Chairman Dan Rostenkowski marked up its bill through the fall of 1985 and the House passed it by voice vote on December 17. The Senate Finance Committee under Chairman Bob Packwood then wrote its own version in the spring of 1986, shifting the design toward the lower rates and broader base that became the final shape, and the Senate passed it 97 to 3 on June 24. The conference committee spent the summer reconciling two genuinely different bills, with the rate structure and the treatment of capital gains among the last issues settled, before the chambers adopted the conference report in September. At every stage the minority participated, offered amendments, and extracted concessions, which is why the final product belonged to both parties.

The 2017 act traveled a compressed and one-sided road. The House Ways and Means Committee released its bill in early November 2017 and marked it up within days; the House passed it 227 to 203 on November 16. The Senate Finance Committee produced its own version, the Senate passed it on December 2, and a conference committee reconciled the two chambers’ bills in less than two weeks, releasing the conference report on December 15. On December 19 the parliamentarian’s Byrd-rule rulings struck the short title and two provisions, the waiver vote failed 51 to 48, and the Senate passed the amended bill 51 to 48 in the early hours of December 20; the House passed it 224 to 201 later that day and the president signed it on December 22. From the first committee markup to signature took roughly six weeks, and at no stage did the majority need or seek a minority vote. The two roads thus embody the procedural axis before the analysis even begins: one built a coalition through prolonged negotiation, the other dispensed with the coalition through a procedure that made it unnecessary.

The Four Axes of the Comparison

The four axes turn two familiar slogans, bipartisan and paid for, into testable claims. Fiscal design compares the budget constraints each statute accepted. Distributional design compares whether each statute held the burden distribution fixed. Procedure compares the legislative vehicles and what those vehicles demanded. Durability compares what survived each statute’s encounter with later Congresses. Each axis carries a practical consequence, because the point of the comparison is not to admire one law and dismiss the other; it is to isolate which choices produced which results, so a future reform can select the mechanism it wants instead of the outcome it happens to remember.

The four-axis comparison table

Comparison axis	Public Law 99-514	Public Law 115-97	Practical consequence of the difference
Fiscal design	Revenue neutral by construction. The JCT estimated about $122 billion in individual revenue loss against about $120 billion in corporate revenue gain over five years, a net of roughly negative $2 billion, so every rate cut had to be financed by repealing a preference or raising revenue elsewhere in the bill.	Enacted through reconciliation under H. Con. Res. 71, section 2001(a), permitting the deficit to increase by not more than $1.5 trillion for the period of fiscal years 2018 through 2027. The JCT estimated the enacted bill would increase the deficit by about $1.46 trillion over that window on a conventional score, excluding macroeconomic feedback.	1986 forced each side to trade preferences for rates and thereby broadened the base. 2017 allowed rate cuts without equivalent offsets and left most preferences in place.
Distributional design	Distributional neutrality treated as a binding design constraint, policed by JCT distributional analysis and only partly achieved in outcome, per the Senate Finance background paper.	No distributional constraint. The Tax Policy Center estimated that in 2018 the top 1 percent received an average cut of $51,000, equal to 3.4 percent of after-tax income, and that by 2027 the top 1 percent would receive 83 percent of the remaining benefit while 53 percent of taxpayers faced a tax increase.	1986’s constraint made a cross-party coalition possible by removing the distributional fight from the negotiation. 2017’s freedom from that constraint shaped both its party-line vote and the distributional debate that followed it.
Procedure	Regular order through both tax-writing committees, a full conference, and final passage with majorities of both parties: House 292 to 136 and Senate 74 to 23 on the conference report, after a 97 to 3 pre-conference Senate vote.	Reconciliation with a 51-vote threshold, no realistic prospect of minority support, and a conference: Senate 51 to 48, House 224 to 201. The short title was struck on a Byrd-rule point of order on December 19, 2017.	Regular order required the minority’s buy-in and spread ownership. Reconciliation enabled speed and a simple-majority threshold but imposed the Byrd rule’s discipline, including the sunset asymmetry.
Durability	The base broadening survived; the rate structure did not. The 28 percent statutory rate, with a 5 percent surcharge producing an effective 33 percent top marginal rate, lasted until the 1990 budget act replaced the surcharge with a 31 percent statutory rate, and the 1993 budget act added 36 and 39.6 percent brackets.	The corporate rate cut to 21 percent is permanent; the individual provisions expire after 2025, except chained-CPI indexing.	Opposite profiles. 1986 changed the base durably and the rates temporarily. 2017 changed the corporate rate durably and the individual provisions temporarily.

The table is the map; the sections below are the territory. Each axis deserves its own treatment because each contains a mechanism that the slogan version of this comparison gets wrong, and the mechanisms interact in ways that matter for the verdict. Fiscal design shaped the coalition, procedure shaped durability, and distributional design shaped both. Readers who work through all four will understand not only that the two statutes differed, but why the differences were forced, not accidental, which is the knowledge that transfers to the next tax debate.

The axes also interact in a specific causal order that the table’s rows do not show. Procedure comes first in time: the majority’s choice of regular order or reconciliation determines whether the minority has leverage. Fiscal design comes second: the budget instruction or the neutrality commitment determines whether rate cuts must be paired with offsets. Distributional design comes third: the presence or absence of a distributional constraint determines whether the bargaining stays on structure or becomes a fight over winners and losers. Durability comes last as the consequence: the coalition the first three axes produced determines what survives. This causal chain is why the article treats the four axes as a system instead of as four separate contrasts. A future reform cannot pick durability à la carte; it must pick the procedure and constraints that produce it, because durability is the output of the chain, not an independent choice. The 1986 act’s durability was manufactured in 1985 and 1986 by the choices the table records; the 2017 act’s impermanence was manufactured the same way.

What the Scores Said Before the Votes

Both statutes were voted on with their official scores published in advance, which means the comparison can be grounded in what Congress knew instead of in what later commentary claimed. The scores are worth walking through provision by provision, because they show exactly what each Congress bought and what it paid.

The 1986 score was a study in balance. The Joint Committee on Taxation estimated the individual income tax provisions would lose about $122 billion over five years while the corporate provisions would gain about $120 billion, for a net revenue effect of roughly negative $2 billion. Read that again: the individual side lost revenue, the corporate side gained it, and the two nearly canceled. The individual rate cuts were the largest losing item, financed on the individual side by the repeal of the investment tax credit’s individual effects, the taxation of capital gains as ordinary income, the passive loss limits, and the repeal of the consumer interest deduction, while the corporate rate cut was more than financed by corporate base broadeners. The score’s message to the voting members was that the bill was a trade, not a gift: lower rates purchased with a broader base, with the purchase price visible in the same table as the goods. That visibility is what made the revenue-neutrality constraint enforceable. Any amendment that deepened a rate cut without adding an offset would have broken the published balance, and the JCT would have said so before the vote.

The 2017 score was a study in authorized imbalance. The JCT’s revenue table for the conference agreement, summarized by CRS in R48485, listed the major items over fiscal years 2018 through 2027. The individual rate cuts lost about $1,214.2 billion. The 20 percent section 199A pass-through deduction lost about $414.5 billion. The increased individual alternative minimum tax exemption lost about $637.1 billion. The corporate rate reduction from 35 to 21 percent lost about $1,348.5 billion, the single largest item. Full expensing for equipment lost about $86.2 billion. Against those losses stood the offsets: the $10,000 SALT cap gained about $670.6 billion, the section 965 transition tax gained about $338.8 billion, the interest deduction limit gained about $253.4 billion, the net operating loss limits gained about $201.1 billion, BEAT gained about $149.6 billion, GILTI gained about $112.4 billion, the repeal of the domestic production activities deduction gained about $98 billion, and chained-CPI indexing gained about $133.6 billion, while FDII lost about $63.7 billion and the foreign dividend exemption lost about $223.6 billion. The bottom line was a deficit increase of about $1.46 trillion on the conventional score. The score’s message to the voting members was the mirror of 1986’s: the bill was a purchase financed by the reconciliation instruction’s $1.5 trillion allowance, with offsets covering only a fraction of the rate cuts.

Two features of the 2017 table deserve emphasis because they illuminate the fiscal-design axis. First, the largest offsets were concentrated in a handful of provisions, the SALT cap and the international package, while the rate cuts were spread across the individual and corporate codes; the bill was not built from matched pairs of cuts and offsets but from a general permission to lose revenue. Second, the table’s totals depended on the sunsets: the individual items were scored only through 2025, so the $1.46 trillion figure measures a bill that was designed to expire in large part. An extension of the individual provisions would have scored far worse, which is why the sunsets were load-bearing for the Byrd rule, not decorative. The 1986 table, by contrast, needed no sunsets because its totals balanced without them.

The scores also shaped the distributional debate before the votes. The JCT’s distributional analysis of the 1986 bill let supporters show, in the official tables, that the burden shares by income class were roughly preserved, which defused the winners-and-losers fight. The JCT’s and the Tax Policy Center’s distributional analyses of the 2017 conference agreement showed the opposite: large early gains concentrated near the top, in the Tax Policy Center’s December 18, 2017 analysis an average 2018 cut of $51,000 for the top 1 percent, equal to 3.4 percent of after-tax income, and a 2027 reversal in which the top 1 percent captured 83 percent of the remaining benefit while 53 percent of taxpayers faced a tax increase. Both sides voted with these tables in front of them. The 1986 tables made a bipartisan vote easier; the 2017 tables made a party-line vote inevitable, because no score existed under which the minority would have accepted the majority’s distributional outcome.

How the scores were used on the floor differed as much as the scores themselves. In 1986, the JCT’s revenue and distributional tables were treated as common facts: both parties cited the same numbers, and disputes were about provisions, not about arithmetic. The neutrality constraints made this possible, because a bill scored as neutral and distributionally neutral gives neither side a numerical grievance to campaign on. Floor debate could therefore focus on the merits of specific base broadeners, whether the passive loss limits were too strict, whether the capital gains repeal went too far, instead of on the bill’s overall fiscal character. In 2017, the scores were treated as contested terrain: the majority cited the JCT’s conventional totals alongside dynamic analyses suggesting larger growth effects, while the minority cited the same JCT tables to emphasize the $1.46 trillion deficit increase and the Tax Policy Center’s distributional findings. The dispute was not about provisions but about which numbers described reality. That difference in floor-debate character follows directly from the fiscal-design axis: a neutral bill produces one shared scoreboard, while a deficit-financed bill produces as many scoreboards as there are models.

The scores also reveal what each Congress considered the relevant time horizon. The 1986 JCT score covered five years, the standard window of the era, and the bill’s permanence meant the five-year picture was representative of the longer run: a neutral bill stays neutral. The 2017 JCT score covered ten years, and the bill’s sunsets meant the ten-year picture was flattering relative to the longer run: a bill whose individual provisions expire looks cheaper over ten years than over twenty, because the out years of the window capture the expiration. Analysts who extended the 2017 provisions in their models produced substantially larger deficit figures, which is why the extension debate is fiscally consequential in a way no 1986 follow-on debate was. The choice of window thus interacted with the durability design: sunsets do not merely satisfy the Byrd rule but also improve the ten-year score, giving drafters a fiscal incentive for impermanence on top of the procedural one. A student of scoring should note this interaction, because it means the 2017 act’s $1.46 trillion figure is best read as the cost of the bill as written, not as the cost of the policy as its supporters hoped it would continue.

Axis One: Fiscal Design

Fiscal design is the axis where the two statutes are most nearly opposites, and it is the axis from which most of the other differences flow. The 1986 act was revenue neutral by construction. That phrase deserves unpacking, because revenue neutrality was not a hope or a forecast; it was the operating rule of the negotiation. The Joint Committee on Taxation estimated the package to be roughly revenue neutral over fiscal years 1987 through 1991, and the political leadership of both parties treated that estimate as a wall. Every rate cut inside the bill had to be financed by a base-broadening provision inside the same bill: repealing the investment tax credit, taxing capital gains as ordinary income, limiting passive losses, repealing consumer interest deductions, or one of the many smaller preference repeals that filled the conference report. The constraint worked like a budget inside the tax-writing committees. Chairmen Dan Rostenkowski in the House and Bob Packwood in the Senate could not buy support with a lower rate unless they simultaneously found the offset, which meant the negotiation was always about trades, not giveaways. Supporters of lower rates had to name the preferences they would sacrifice, and defenders of preferences had to name the rate cuts they would forgo. That discipline is what made the base broadening real, not rhetorical. It also explains why the 1986 base changes survived: provisions repealed to pay for a rate cut are repealed by a coalition that includes the rate cut’s beneficiaries, which makes them harder to restore later than provisions repealed by fiat.

The 2017 act worked under the opposite fiscal design. It moved through reconciliation under H. Con. Res. 71, whose section 2001(a) permitted the deficit to increase by not more than $1.5 trillion for the period of fiscal years 2018 through 2027, which allowed the bill to pass the Senate with 51 votes. The instruction was not a ceiling the drafters chafed against; it was the resource they spent. The Joint Committee on Taxation estimated in JCX-67-17 that the conference agreement would increase the deficit by about $1.46 trillion over that window on a conventional score, excluding macroeconomic feedback, landing just inside the permitted loss. Because the budget instruction allowed a large revenue loss, rate cuts did not have to be matched by preference repeals, and for the most part they were not. The 2017 act did include some base broadeners, notably the $10,000 cap on the state and local tax deduction, the suspension of the personal exemption, and the repeal of the domestic production activities deduction, but these were modest beside the scale of the rate cuts, and the statute’s largest base-related move, the redesigned international system, was a restructuring of how foreign earnings were taxed instead of a repeal of domestic preferences. The fiscal design thus explains the 2017 act’s shape: a large corporate rate cut, broad individual rate relief, and a base that was trimmed at the margins instead of rebuilt from the foundation.

The practical consequence of the fiscal-design difference is the one the table states and the one the verdict will rest on. Revenue neutrality forced the 1986 coalition to trade preferences for rates, and the trades broadened the base. The reconciliation instruction freed the 2017 coalition from trading, and the base stayed largely as it was. Neither choice was dictated by economics; each was dictated by the budget procedure the majority selected, and by how much deficit increase the majority was willing to absorb. A future reform that wants the 1986 result, a broader base, must therefore accept the 1986 constraint, because base broadening is what happens when rate cuts have to be paid for, not what happens when they do not.

The reconciliation instruction deserves a closer look as a fiscal-design instrument, because it functioned less like a limit and more like an appropriation. The $1.5 trillion figure in H. Con. Res. 71, section 2001(a), was negotiated among the majority as the amount of deficit increase the party could tolerate, and the tax-writing committees then treated it as a budget to spend. Committee markups are revealing in this regard: provisions were added, enlarged, or trimmed with an eye on the running JCT total, and the conference agreement’s final score of about $1.46 trillion shows the drafters spending nearly the full allowance while keeping a small margin for estimation error. This is the normal behavior of a reconciliation instruction. It does not merely permit a deficit increase; it invites the majority to allocate one, and the allocation becomes the bill’s real fiscal policy. The 2017 act’s shape, a large corporate cut, broad individual relief, and modest offsets, is what $1.5 trillion of permitted deficit increase buys when the buyers want rate cuts.

Contrast the 1986 committees’ behavior under revenue neutrality. There the running JCT total had to stay near zero, so every addition to the bill required a subtraction. The conference committee’s work in the summer of 1986 was dominated by this arithmetic: when the Senate’s lower rates threatened to break the balance, the conferees found additional base broadeners; when a base broadener proved politically impossible, the rates moved up. The discipline was continuous because the score was continuous, published and updated as the bill changed. The two committees thus behaved as mirror images: one spending a deficit allowance down to its floor, the other defending a zero balance against every amendment. A student of fiscal design should internalize this symmetry, because it explains why the two statutes’ contents look the way they do far better than any account centered on ideology.

There is also a temporal dimension to the fiscal-design difference that matters for the durability axis. The 1986 act’s offsets were permanent repeals, so its fiscal balance did not depend on future Congresses doing anything; the base stayed broad because the repealed provisions stayed repealed. The 2017 act’s fiscal balance, such as it was, depended on the sunsets: the $1.46 trillion score assumed the individual provisions would actually expire, and any extension would worsen the fiscal outcome without any new vote on the original bill’s design. The 2017 fiscal design thus contained a time bomb that the 1986 design did not: a future Congress that extends the individual provisions must either find new offsets or accept a larger deficit increase than the original score showed. The fiscal choice of 2017 was therefore not a one-time event but a contingent one, whose full cost depends on what later Congresses do with the expirations.

The 1986 conference’s rate-for-base trades show the fiscal-design mechanism operating at full intensity. The Senate bill had pushed individual rates lower than the House bill, and the conference had to find the offsets to pay for the difference without breaking either the revenue-neutrality or the distributional-neutrality constraints. The conferees’ solution combined additional base broadeners with the 5 percent surcharge that produced the effective 33 percent top marginal rate: the surcharge raised revenue from high-income filers while preserving the 28 percent statutory rate as the bill’s headline, which kept the distributional tables in balance. Every percentage point of rate cut in that negotiation had a named pay-for attached, and the conference report’s JCT tables let every member see the trade. Contrast the 2017 conference, whose work was dominated by a different arithmetic: allocating the $1.5 trillion instruction across the chambers’ competing priorities, with the corporate rate’s depth, the individual rate cuts’ size, and the SALT cap’s level as the main variables. The 2017 conferees traded priorities against the deficit allowance instead of trading rates against preferences, which is why their product contained so few matched pairs. The two conferences were both called conferences, but they performed opposite fiscal operations: one balanced a scale, the other spent a budget.

Why did revenue neutrality force the 1986 coalition to repeal preferences?

Revenue neutrality turned every rate cut into a purchase that had to be paid for inside the bill, so supporters of lower rates had to name the preferences they would sacrifice. That enforced trading converted rate-cutting ambition into actual base broadening, and it gave each repealed preference a cross-party set of defenders against later restoration.

There is a further fiscal-design subtlety worth stating because it recurs in every later debate about paying for tax cuts. The 1986 act’s revenue neutrality was measured on a static basis by the Joint Committee on Taxation, meaning the estimate did not assume that lower rates would generate enough additional growth to replace the repealed preferences. Later advocates of unpaid-for rate cuts sometimes argued that the 1986 experience showed rate cuts paying for themselves through growth; the record shows the opposite, which is that the 1986 rate cuts were paid for by identified offsets, about $122 billion of individual revenue loss matched against about $120 billion of corporate revenue gain, and scored as essentially neutral without any growth assumption. The 2017 act, by contrast, was accompanied by claims that faster growth would close much of the revenue gap, and the conventional score of about $1.46 trillion in added deficits over fiscal years 2018 through 2027, consistent across the JCT and CBO conventional estimates, records the gap the sponsors’ arithmetic left open. This article takes no position on the deeper growth debate; it records only that the two statutes made opposite wagers about financing, that the wagers were visible in the scores published before the votes, and that the scores, not the rhetoric, are what a student of fiscal design should study. The lesson for a future reform is procedural, not ideological: the budget constraint a bill accepts determines the coalition the bill needs, and the coalition determines what the bill can change.

Axis Two: Distributional Design

Distributional design is the quieter axis and, by the brief’s own framing, the source of most of the difference in the two statutes’ politics. The 1986 act treated distributional neutrality as a binding constraint in design. In practice that meant the tax-writing committees required the bill to leave the distribution of the income tax burden across income classes roughly unchanged: the shares of total taxes paid by low, middle, and high-income filers after the bill were supposed to look like the shares before it. The Joint Committee on Taxation’s distributional analysis policed the constraint, and the leadership treated a violation the way it treated a revenue violation, as a reason to renegotiate the offending provision. The Senate Finance Committee’s background paper framed neutrality as a design requirement, not a forecast, and the record shows it succeeded as a negotiating rule more completely than as a measured result: later analysis found the constraint only partly achieved in outcome, with the final distribution shifting at the margins even as the negotiation honored the rule. The constraint did not express a theory of fairness; it expressed a theory of coalition management. As long as no income class could be shown to be a systematic winner or loser, neither party’s distributional constituencies had a clean reason to defect, and the negotiation could stay focused on the structure of the code instead of on who would pay for it. The repeal of preferences and the cut in rates were designed to offset each other within each income group, which is why the statute could combine a large rate cut at the top with the removal of millions of low-income filers from the rolls without triggering a distributional revolt.

The mechanism deserves a closer look, because distributional neutrality is often misread as a claim that the 1986 act left everyone untouched. It did not. Within each income class there were winners and losers: a high-income filer who had relied on the investment tax credit or passive losses could face a higher effective burden even as the top statutory rate fell, while a high-income filer with plain wage income enjoyed the full rate cut. What the constraint was designed to hold constant was the class-level aggregate, not the individual outcome, and the record shows it succeeded as a negotiating rule more completely than as a measured result. That distinction is what made the constraint politically useful. It allowed each party to tell its constituencies that the bill as a whole did not redistribute the burden, while the internal reshuffling of who paid what was buried in the complexity of base-broadening provisions that few voters tracked provision by provision. The 1986 coalition thus bought its supermajority partly with opacity: the distributional constraint was enforced at a level of aggregation where compliance was verifiable by the JCT and invisible to the public.

The 2017 act accepted no distributional constraint, and the difference shows in both the estimates published around passage and the politics that followed. The Tax Policy Center, in its December 18, 2017 distributional analysis of the conference agreement, estimated that in 2018, the first full year under the new law, the top 1 percent of filers received an average tax cut of $51,000, equal to 3.4 percent of after-tax income. The Joint Committee on Taxation’s own distributional analysis of the conference agreement, published before the final votes, showed average tax changes varying by income class and year in ways that reflected the mix of permanent corporate relief and temporary individual relief. Looking further out, the Tax Policy Center estimated that by 2027, after the individual provisions expired, the top 1 percent would receive 83 percent of the total remaining benefit while roughly 53 percent of taxpayers would face a tax increase relative to the law as it stood before the 2017 act. The 2027 result follows mechanically from the sunset design: the corporate rate cut and the international provisions persist, while the individual rate cuts, the doubled standard deduction, and the section 199A deduction disappear, so the remaining law raises more from individuals than the prior baseline did. None of these estimates was hidden; all were published before the final votes. The point is not that the estimates were secret but that no rule required the bill to satisfy them.

The politics followed the design. Because the 2017 act left distribution free to move, the distributional question became the central public fight over the bill: who benefits, by how much, and for how long. Supporters emphasized the 2018 average gains and the corporate provisions’ investment incentives; opponents emphasized the concentration of gains at the top and the 2027 reversal for individual filers. That fight is exactly what the 1986 constraint had been designed to prevent, and its absence is a large part of why the 2017 act passed on party lines. When a bill’s distributional effects are unconstrained, each side can recruit the estimates that favor its case, the minority has no reason to accept the majority’s numbers, and the negotiation over structure becomes a negotiation over winners and losers, which is a negotiation the minority cannot win and therefore will not join.

Why did distributional neutrality make a bipartisan coalition possible in 1986?

Distributional neutrality removed the winners-and-losers fight from the negotiation by requiring the JCT-scored burden shares by income class to stay roughly constant. With distribution held fixed, neither party’s constituencies had a clean reason to defect, so the bargaining stayed on structure, rates against preferences, where trades across party lines were possible.

A final distributional subtlety concerns the corporate side, where the two statutes’ designs diverged as sharply as on the individual side. The 1986 act cut the corporate rate from 46 to 34 percent but paid for much of that cut with corporate base broadeners, including the repeal of the investment tax credit, so the corporate sector as a whole financed its own rate relief. The 2017 act cut the corporate rate from 35 to 21 percent permanently without equivalent corporate offsets; the Joint Committee on Taxation estimated the corporate rate reduction alone at a revenue loss of about $1,348.5 billion over the ten-year window. The distributional consequence of an unpaid-for corporate rate cut depends on contested assumptions about who bears the corporate tax, with the JCT and the Tax Policy Center using different incidence assumptions that produce different distributions, and this article does not adjudicate between them. What matters for the comparison is the design difference: 1986 paired its corporate rate cut with corporate base broadening under a binding neutrality constraint, while 2017 paired its larger corporate rate cut with no such pairing and no such constraint. The distributional debate that followed each statute was therefore about different things: in 1986, about whether the JCT’s class-level neutrality held up in practice, and in 2017, about the direction and magnitude of redistribution itself.

The methodological fine print behind the 2017 distributional estimates deserves attention, because it explains why the two sides could cite different numbers without either side inventing them. Distributional analysis requires assumptions about tax incidence, meaning who ultimately bears the burden of each tax. For the individual income tax the incidence is straightforward: the filer bears it. For the corporate income tax it is contested: the JCT has traditionally assigned a large share of the corporate burden to owners of capital, while the Tax Policy Center’s model distributes it across capital owners and workers in proportions that shift over time. Because the 2017 act’s largest permanent provision was the corporate rate cut, the choice of incidence assumption materially changes the distributional picture: models that assign more of the corporate tax to workers show the rate cut benefiting workers more, while models that assign it to shareholders show the benefit concentrating at the top. The Tax Policy Center’s headline figures, the $51,000 average 2018 cut for the top 1 percent and the 83 percent share of the 2027 remaining benefit, reflect its incidence choices, and the JCT’s tables reflect its own. Neither is fraudulent; they answer slightly different questions with slightly different tools.

This methodological pluralism is itself a product of the missing constraint. Under the 1986 regime, the JCT’s distributional tables were the single official scoreboard, and the negotiation optimized against them; alternative models existed but did not govern. Under the 2017 regime, with no official distributional constraint, every model became ammunition: supporters cited analyses showing broad-based gains in the early years, opponents cited analyses showing concentration at the top and the 2027 reversal, and the public debate became a contest of models, not a negotiation over a shared table. The lesson for a future reform is that distributional neutrality does double duty: it constrains the bill’s substance and it designates the scoreboard, depriving the debate of the model-shopping that characterized the 2017 fight. A Congress that wants a quiet distributional debate should therefore want a binding distributional constraint, whatever its views on the right distribution.

One more distributional contrast clarifies the stakes. The 1986 act’s largest low-income provision, the combination of the nearly doubled standard deduction, the enlarged personal exemption, and the expanded earned income tax credit, removed millions of filers from the rolls permanently and was financed inside the revenue-neutral package. The 2017 act’s low-income provisions, the doubled standard deduction and the expanded child credit, were larger in nominal dollars for many families but temporary, expiring with the rest of the individual title after 2025. The distributional design difference thus reproduced itself at the bottom of the income scale as well as the top: 1986 bought permanent relief for low-income filers with permanent offsets, while 2017 rented temporary relief with a temporary deficit increase. Which arrangement a reader prefers is a question this article does not answer; that the arrangements differ structurally is the point.

The earned income tax credit expansion deserves emphasis as the 1986 distributional story’s positive side, since the neutrality constraint is sometimes misread as purely defensive. The credit’s expansion was one of the largest anti-poverty measures enacted through the tax code up to that point, and it was financed inside the same revenue-neutral bargain as the rate cuts: the JCT’s tables showed the package holding the burden distribution steady in part because the credit moved resources down the income scale while the base broadening moved the tax burden up it. The design thus paired a progressive spending-like provision with regressive-leaning base changes, and the neutrality constraint is what forced the pairing. Without the constraint, the rate-cutting coalition would have had no reason to include the credit expansion; with it, the expansion became the price of the coalition’s distributional balance. This is the neutrality price operating in miniature: the constraint forced each side to name what it would give up, and what the rate cutters gave up was revenue that financed relief for filers at the bottom. The 2017 act’s expanded child credit played a comparable political role, sweetening the individual title for middle-income families, but without a neutrality constraint to anchor it, the credit expires with everything else after 2025.

Axis Three: Procedure

Procedure is the axis where the two statutes’ differences were most visible to the public and most misunderstood by it. The 1986 act moved under regular order, which in tax legislation means the bill originated in the House Ways and Means Committee, received a markup, passed the House, moved to the Senate Finance Committee for its own markup, passed the Senate in amended form, and went to a conference committee that reconciled the two versions before each chamber voted on the conference report. Regular order is slow, permeable, and demanding: at every stage the minority can offer amendments, force votes, and slow the bill, which means the majority must either win minority votes or accept a bill the minority can live with. The 1986 process used every stage. The House passed its version by voice vote on December 17, 1985, after a Ways and Means markup under Chairman Rostenkowski. The Senate Finance Committee produced its own version under Chairman Packwood in the spring of 1986, the Senate passed it 97 to 3 on June 24, and the conference committee spent the summer reconciling two genuinely different bills before the House adopted the conference report 292 to 136 on September 25 and the Senate adopted it 74 to 23 on September 27. The mechanics of that route, and why it matters for what a bill can contain, are explained in how tax bills move through Congress.

The 2017 act moved through reconciliation, which is a different legislative universe with different rules and different consequences. Reconciliation begins with a budget resolution that gives committees instructions: raise or lower revenue, spending, or deficits by specified amounts within a specified window. A bill written to those instructions can pass the Senate with 51 votes, bypassing the filibuster, but it must satisfy the Byrd rule, which bars provisions whose budgetary effects are merely incidental to non-budgetary policy and, critically, bars any provision that increases the deficit beyond the budget window, typically ten years. The fiscal year 2018 budget resolution gave the tax-writing committees reconciliation instructions permitting up to $1.5 trillion in added deficits over fiscal years 2018 through 2027. The House Ways and Means Committee marked up its bill in November 2017, the House passed it on November 16, the Senate Finance Committee produced its version, the Senate passed it on December 2, a conference committee reconciled the two, and the conference report passed both chambers on December 20 before signature on December 22. The calendar is the tell: from committee markup to presidential signature in roughly six weeks, a pace regular order could not have sustained.

The procedural difference produced the sunset asymmetry, which is the single most misunderstood feature of the 2017 act. Because the Byrd rule forbids a reconciliation bill from increasing the deficit beyond the ten-year budget window, any provision that loses revenue in year eleven or later must either be offset or be given an expiration date inside the window. The drafters chose expiration. The corporate rate cut to 21 percent was scored as losing revenue beyond the window too, but the instruction’s $1.5 trillion allowance covered the window years, and the permanent corporate cut was structured to fit within the Byrd rule’s requirements as scored, while the individual provisions, the doubled standard deduction, the lower individual rates, the section 199A deduction, the suspended personal exemption, the SALT cap, were all given a December 31, 2025 expiration. That is why the statute has two clocks: a permanent corporate clock and a temporary individual clock. The 1986 act, moving under regular order, faced no Byrd-rule constraint and therefore had no sunsets; its provisions were permanent law until later Congresses changed them, which they did, but through new legislation instead of through expirations written into the original bill.

Why did the 1986 act take nearly two years while the 2017 act took six weeks?

Regular order gave the minority amendment rights and filibuster leverage at every stage, so the 1986 majority had to negotiate continuously from the 1985 House markup through the 1986 conference. Reconciliation’s 51-vote threshold removed that leverage, letting the 2017 majority move from markup to signature in about six weeks without needing a single minority vote.

Procedure also shaped the coalitions in a way the vote counts make plain. Regular order’s minority leverage meant the 1986 bill could not pass without minority support, so the drafters built the support in: the conference report drew votes from both parties in both chambers, culminating in the 292 to 136 House and 74 to 23 Senate votes on the conference report. The detailed account of how that coalition was assembled, committee by committee and trade by trade, is told in the passage history of the 1986 act. Reconciliation’s 51-vote threshold meant the 2017 bill did not need minority support, so the drafters did not build it: the final votes were 51 to 48 in the Senate and 224 to 201 in the House, with every minority member opposed. The procedural choice thus determined the coalition choice. A majority that can pass a bill alone will generally pass it alone, because building a bipartisan coalition costs concessions, and a majority that must have minority votes will generally pay for them, because the alternative is no bill at all. The 1986 supermajority and the 2017 party-line vote were not accidents of personality or of the political climate; they were the predictable outputs of the procedures each majority selected.

Why do the 2017 individual provisions expire after 2025?

The Byrd rule bars a reconciliation bill from increasing the deficit beyond the ten-year budget window, so revenue-losing provisions that extend past the window must expire inside it. The drafters gave the individual provisions a December 31, 2025 sunset to satisfy the rule, while structuring the corporate rate cut to comply without expiring.

There is a temptation to treat the 2017 procedure as a shortcut that achieved the same destination faster, and the sunset asymmetry is the refutation. A reconciliation bill is not a regular-order bill passed quickly; it is a different kind of law, one whose content is disciplined by the Byrd rule instead of by the minority’s amendments. The Byrd rule struck the 2017 bill’s short title, forced the sunsets, and shaped which provisions could be permanent. Regular order imposes no such discipline, which is why the 1986 act could be revenue neutral, distributionally neutral, and permanent all at once: the discipline came from the negotiation instead of from the parliamentarian. Each procedure thus carries a characteristic cost. Regular order costs time and concessions. Reconciliation costs permanence for anything the budget window cannot absorb and costs the minority’s ownership of the result. A future reform must choose its procedure knowing which cost it is willing to pay, because the procedure will write large parts of the bill before the policy debate even begins.

The December 2017 endgame illustrates reconciliation’s characteristic cost in miniature. The conference report released on December 15 contained three provisions the parliamentarian ruled extraneous under the Byrd rule: the short title itself, the 529-account expansion for home-schooling expenses, and the endowment-tax exception for colleges with fewer than 500 tuition-paying students. The Senate struck them on December 19, the waiver vote failed 51 to 48, and the bill the Senate passed in the early hours of December 20 was therefore textually different from the bill the House had passed the previous afternoon. The House had to vote again, which it did 224 to 201, and only then could the bill go to the president. The episode is sometimes told as a comic footnote, the bill so rushed that it lost its own name, but its structural meaning is serious: the Byrd rule is not a formality the majority can waive at will, since waiver requires 60 votes the reconciliation majority by definition lacks, and every reconciliation bill is therefore written in negotiation with the parliamentarian as a silent conferee. The 2017 act’s final shape, including its sunsets, was co-authored by that silent conferee.

The 1986 conference worked under no such shadow, and its amendment culture shows the difference. Because regular order gave the minority procedural rights, the 1986 bill accumulated bipartisan fingerprints at every stage: the House bill reflected Rostenkowski’s negotiations with Republicans, the Senate bill reflected Packwood’s, and the conference reconciled provisions that each chamber’s minority had helped shape. The famous 97 to 3 pre-conference Senate vote was not an accident of an easy bill; it was the product of a Finance Committee markup in which the minority had been given real victories to defend. By the time the conference report reached the floor, dozens of senators had provisions in the bill they had personally negotiated, which is why the final 74 to 23 vote, though less lopsided than the pre-conference tally, still drew majorities of both parties. Contrast the 2017 conference, whose members were drawn from the majority alone and whose product no minority member had reason to defend. The vote counts measured the procedures with precision: a bill the minority helped write gets minority votes, and a bill written without the minority does not.

Procedure also determined what each bill could not do. Regular order could not have produced the 2017 act’s deficit increase, because the minority would have filibustered a $1.46 trillion unpaid-for tax cut and the majority could not have bought the votes without offsets. Reconciliation could not have produced the 1986 act’s base broadening, because the majority writing alone would never have repealed its own donors’ preferences without the cover of a bipartisan bargain; the investment tax credit died in 1986 precisely because both parties’ fingerprints were on the knife. Each procedure thus enables a different kind of substance, and the substance each enabled is the substance each produced. This is the deepest procedural lesson of the comparison: the vehicle is not neutral packaging around a policy choice but a constraint that selects which policies are writable.

The Parliamentarian’s Role in 2017

No comparison of the two procedures is complete without the Senate parliamentarian, whose rulings under the Byrd rule shaped the 2017 act more directly than any single legislator’s preferences. The Byrd rule, named for Senator Robert Byrd, bars extraneous matter from reconciliation bills. A provision is extraneous if it does not change outlays or revenues, if it increases the deficit beyond the budget window, or if it falls outside the jurisdiction of the committee that reported it, among other tests. Sixty votes can waive the rule, but a majority using reconciliation to avoid the filibuster rarely has sixty votes, which is why the rule binds.

The 2017 drafters wrote the bill in the shadow of these tests from the beginning. Every provision had to produce a budgetary effect within the fiscal year 2018 to 2027 window, and no provision could increase the deficit beyond that window. The individual sunsets after 2025 are the most consequential product of this shadow: permanent individual cuts would have failed the beyond the window test, so the drafters wrote expiration dates instead. The parliamentarian’s review, conducted provision by provision in the days before final passage, struck or forced changes to items that failed the tests. The short title’s removal on December 19, 2017, after Senator Bernie Sanders raised the point of order and Senator Mike Enzi’s waiver failed 51 to 48, was the most visible ruling, but the substantive rulings on the bill’s provisions mattered more for the statute’s shape.

Contrast the parliamentarian’s role with the role of the conference committee in 1986. The 1986 conference reconciled substantive differences between two bills that had each survived committee scrutiny, and its decisions were political judgments about which preferences would die and which rates would fall. The 2017 process added a second, procedural conference running in parallel: the negotiation with the parliamentarian over what reconciliation could carry. The 1986 bill was shaped by bargaining between the parties. The 2017 bill was shaped by bargaining between the majority and the rules. Both are forms of negotiation, but they produce different kinds of legislation, because the parliamentarian enforces procedure while a conference enforces compromise.

The practical consequence is that reconciliation does not merely speed up legislation. It changes what legislation can contain. Permanent structural change is possible, as the corporate rate shows, but only for provisions whose cost fits the cap and whose effects satisfy the Byrd tests. Everything else must sunset, phase, or be dropped. The 2017 act is therefore not just a policy choice but a demonstration of the reconciliation procedure’s substantive limits, and any future majority considering the same path should study those limits before assuming it can replicate the 1986 scope without the 1986 constraints.

What Regular Order Cost and Bought

Regular order is expensive, and the 1986 price tag is worth itemizing, because the cost is the reason later majorities stopped paying it. The process consumed nearly two years from the Treasury proposals to the presidential signature. It nearly died in the Senate Finance Committee in the spring of 1986, when the markup stalled and the bill’s survival was in genuine doubt. It required the House Ways and Means chairman to spend months holding a coalition that disagreed on almost everything except the framework. It forced both parties to vote for the elimination of preferences their own donors and constituents valued. And it required the president to accept a bill that raised corporate revenue to pay for individual rate cuts, a trade many in his own party opposed.

Each of those costs was also a purchase. The two years bought the hearings and markups that exposed every provision to scrutiny, which meant the final bill contained fewer surprises and fewer drafting errors than a fast bill could have managed. The near death in Finance bought the deeper bargain: Packwood’s revival through steeper rate cuts financed by steeper base broadening only worked because the committee had to find a package that could command a majority under the neutrality rules, and the package it found was more ambitious than the one that had stalled. The months Rostenkowski spent holding the House coalition bought the voice vote passage in December 1985, a signal of breadth that strengthened the bill in the Senate. The painful preference eliminations bought the rate cuts themselves, which was the only currency the rules accepted. And the president’s acceptance of the corporate revenue increase bought the individual rate cuts that were the bill’s political purpose.

The purchase that matters most for the comparison is legitimacy across party lines. A bill that both parties helped write is a bill both parties have reasons to defend, and the 1986 base provisions survived the 1990 and 1993 budget bills in part because Democratic and Republican legislators alike had voted for them and understood them. Contrast the position of a minority party that voted unanimously against a bill: it has every incentive to let the bill’s temporary provisions expire, to attack its distributional effects, and to deny it any legitimacy beyond the votes that passed it. Regular order buys the minority’s fingerprints on the legislation. Reconciliation buys speed and scale. The 1986 Congress paid the first price. The 2017 Congress paid the second.

There is a further cost of regular order that deserves honesty: it empowers obstruction. Every stage that gives the minority leverage to enforce neutrality also gives the minority leverage to kill the bill, and the 1986 process came close to that outcome more than once. A majority that wants guaranteed results may rationally prefer reconciliation, which removes the minority’s veto at the cost of removing the minority’s participation. The comparison does not hide this tradeoff. It presents it as the central institutional choice: regular order risks failure and buys durability, while reconciliation guarantees passage and buys fragility. The two statutes are the receipts for the two purchases.

The Minority’s Leverage

The procedural axis of this comparison can be restated as a question about leverage: what power does the minority party hold in each process, and what does it do with that power. The answer explains both the coalitions and the durability profiles, because leverage determines who must be accommodated and therefore whose fingerprints end up on the final bill.

Under regular order, the minority’s leverage is structural and continuous. In committee, the minority can offer amendments, force recorded votes, and use the markup process to extract concessions provision by provision. On the floor, the minority can debate, amend within the rules of each chamber, and use the Senate’s extended procedures to slow or block legislation. In conference, the minority’s conferees participate in reconciling the two chambers’ bills, which gives them influence over the final text. At every stage, the majority needs the minority’s votes or at least its acquiescence, and that need is what converts the minority from spectators into negotiators. The 1986 minority used this leverage to enforce the distributional neutrality constraint: Democratic legislators in both chambers could credibly threaten to walk away, and the threat is what kept the bill from becoming a vehicle for shifting the burden toward their constituents.

The leverage also shaped the substance in ways the vote totals conceal. The 1986 conference report’s 292 to 136 House vote and 74 to 23 Senate vote look like comfortable margins, but the margins were built from dozens of small accommodations made under the pressure of minority leverage. The expanded standard deduction and personal exemption, which protected low income households, were minority demands. The calibration of the base broadening against the rate cuts, which held the distribution steady, was a product of minority scrutiny in committee. The surcharge that created the effective 33 percent top rate was a technical solution to a distributional problem the minority had identified. None of these provisions would have existed in the same form without a minority empowered to insist on them, which is why the bill’s durability owes as much to its opponents’ leverage as to its supporters’ vision.

Under reconciliation, the minority’s leverage collapses to a single point: the Byrd rule. The minority cannot filibuster, cannot amend freely on the floor, and has no conferees in any meaningful sense. Its only procedural weapon is the point of order against extraneous provisions, raised before the parliamentarian, which can strip individual items but cannot reshape the bill’s design. The 2017 minority used this weapon where it could. Senator Bernie Sanders’s point of order against the short title, sustained when Senator Mike Enzi’s waiver failed 51 to 48, was the most visible use, but the Byrd bath stripped or modified substantive provisions as well. These were tactical victories within a strategic defeat: the minority could remove the bill’s name but not its rate cuts, could trim provisions but not impose neutrality.

The minority’s remaining leverage under reconciliation is political instead of procedural. It can vote no unanimously, as the Democrats did in both chambers in December 2017, and use that unanimity as the foundation for future attacks on the bill’s distributional effects and its sunsets. It can promise to let the temporary provisions expire, which converts the majority’s sunset strategy into a future negotiation the minority may win. And it can deny the bill the legitimacy that bipartisan support confers, which matters for durability even when it does not matter for passage. The 2017 minority did all three, and the scheduled expiration of the individual provisions after 2025 is the arena in which that leverage will be tested.

The comparison thus closes where it began, with the constraints. Regular order gives the minority leverage, leverage enforces neutrality, neutrality builds the coalition, and the coalition produces durability. Reconciliation removes the leverage, removes the neutrality, narrows the coalition, and schedules the expiration. The minority’s leverage is not a footnote to the four axes. It is the transmission belt that connects procedure to durability, and the two statutes show the belt engaged in one case and disengaged in the other.

Axis Four: Durability

Durability is where the comparison pays off, because the two statutes have opposite durability profiles and the opposition reveals what actually makes tax law stick. The political economy of repeal explains most of it. Repealing a preference is an act of destruction whose political costs are paid once, by the coalition that votes for the repeal, while restoring a preference is an act of creation whose costs must be paid anew: a sponsor must be found, a JCT score must be absorbed, and a coalition must be assembled around a provision whose beneficiaries have spent the intervening years adapting to its absence. The investment tax credit illustrates the asymmetry. Its repeal in 1986 was financed by the rate cuts it made possible, so the repeal coalition included everyone who wanted lower rates. Its restoration would have to be financed by a rate increase or a new deficit, so no coalition for restoration ever formed, and the credit stayed dead. Every major 1986 base broadener benefited from the same asymmetry, which is why the base survived even as the rates fell.

The durability of the 1986 base also benefited from a subtler mechanism: the repealed preferences stopped generating the lobbying infrastructure that defends live preferences. A live preference has a trade association, a set of congressional champions, and a JCT score that quantifies its cost; a repealed preference has none of these, because its defenders disband once the fight is lost and its champions move on to live issues. The passive loss limits codified at section 469 show the mechanism at work: the syndication industry that had thrived on passive losses reorganized around the new rules within a few years, and by the time anyone might have proposed repeal, the industry’s business models no longer depended on the old preference. Base broadening thus exhibits a ratchet effect. Each repeal destroys the constituency for restoration, which makes the next repeal easier and restoration progressively harder. The 1986 act turned this ratchet through dozens of provisions at once, which is why its base changes proved so resistant to reversal.

Rate changes exhibit the opposite political economy, which is why both statutes’ rate structures proved vulnerable in different ways. A rate cut’s beneficiaries remain organized and visible: every corporation paying 21 percent knows exactly what the 2017 act did for it, and every filer in the 37 percent bracket knows what the individual rate cuts are worth. That visibility makes rate cuts easy to defend in the short run but also easy to target: a future Congress that needs revenue can raise a rate with a single number change, without reopening any of the bargains that produced the base. The 1990 and 1993 acts demonstrated the technique on the 1986 rates, raising the top individual rate in two steps while leaving the base untouched. The technique works because rate increases, unlike preference restorations, require no new coalition for a specific provision; they require only a majority that wants revenue, which deficit pressure reliably supplies. Rates are therefore the shock absorbers of the tax code: the first thing adjusted when fiscal conditions change, and the least durable part of any reform.

The 2017 act’s durability profile follows directly from these political economies, modified by the sunsets. The permanent corporate rate cut enjoys the standard protection of an organized beneficiary class: corporations paying 21 percent will resist any increase, and the resistance does not depend on the enacting coalition’s survival. The expiring individual provisions, however, face a harsher test than ordinary rate cuts, because their beneficiaries must win an affirmative legislative fight to keep them. An ordinary rate cut persists until a majority votes to raise it; a sunsetting provision expires unless a majority votes to extend it. The burden of legislative action is reversed, and reversed burdens favor expiration when Congress is divided or when the extension’s JCT score cannot be financed. The 2017 drafters accepted this reversal knowingly: the sunsets were the price of fitting the bill inside the reconciliation instruction, and the price is paid by the individual provisions’ beneficiaries in the form of permanent political vulnerability.

Start with 1986. Its base broadening survived; its rates did not. The repealed investment tax credit stayed repealed. The passive loss limits, the at-risk rules, the repeal of the consumer interest deduction, and the taxation of capital gains as ordinary income all endured as structural features of the code, and later Congresses built on them instead of reversing them. The passive activity loss limits added by section 502 of the 1986 act, codified at section 469 of the Internal Revenue Code, remain current law, as IRS Publication 925 confirms. The rate structure, by contrast, lasted barely four years. The Omnibus Budget Reconciliation Act of 1990, Public Law 101-508, replaced the 5 percent surcharge with a 31 percent statutory rate, and the Omnibus Budget Reconciliation Act of 1993, Public Law 103-66, added 36 percent and 39.6 percent brackets, per CRS RL34498. The 28 percent statutory rate, with its surcharge producing an effective 33 percent top marginal rate, was gone by the early 1990s, while the base-broadening provisions that had paid for it remained. The pattern is instructive: what 1986 changed durably was the definition of taxable income, and what it changed temporarily was the rate applied to it.

Turn to 2017, whose profile is the mirror image. Its corporate rate cut is permanent; its individual provisions are scheduled to expire. The 21 percent corporate rate has no sunset date and persists until a future Congress affirmatively raises it, which is the same legal posture the 1986 base broadeners occupied: permanent law, changeable only by new legislation. The individual provisions, by contrast, carry their own expiration. After December 31, 2025, the individual rates revert to the pre-2018 schedule, the standard deduction falls back, the personal exemption suspension ends, the SALT cap disappears, and the section 199A deduction ends, unless Congress extends them. The individual side was not uniformly temporary: the shift to chained CPI for indexing brackets and thresholds carries no sunset, while 100 percent expensing for equipment began phasing down in 2022 and phases out by 2026, and the GILTI, FDII, and BEAT provisions step up after 2025. The 2017 act thus changed rates durably on the corporate side and temporarily on the individual side, while changing the base in ways that were real but narrower than 1986’s, through the international redesign and the SALT cap instead of through a wholesale repeal of domestic preferences.

Which statute’s rate structure proved more durable?

The 1986 individual rate structure lasted about four years before the 1990 and 1993 budget acts restored higher brackets. The 2017 corporate rate cut to 21 percent carries no expiration and persists until Congress affirmatively changes it, showing that permanence in the text outlasts permanence in the coalition.

Why did base broadening outlast rate cutting in 1986? The fiscal-design mechanism supplies the answer. Base-broadening provisions repealed to pay for a rate cut are defended, after passage, by the coalition that wanted the rate cut: restoring the preference would require finding new revenue or accepting a higher rate, and neither side of the original bargain wants to reopen it. Rate cuts, by contrast, are defended only by the coalition that wants low rates, and a later Congress facing a deficit can raise rates without disturbing the base bargain, which is exactly what the 1990 and 1993 acts did. They raised the top rates while leaving the broadened base in place, effectively keeping the 1986 structure and repricing it. The durability lesson of 1986 is therefore that base changes are stickier than rate changes, because base changes rearrange interests while rate changes merely reprice them.

The 2017 durability story inverts the mechanism. The permanent corporate rate cut persists not because a cross-party coalition defends it but because permanence was written into the statute and repeal requires affirmative legislation that no later majority has yet assembled. The expiring individual provisions, meanwhile, persist only if Congress acts to extend them, which reverses the usual legislative burden: instead of needing a majority to change the law, the law changes itself unless a majority intervenes. That reversal is the sunset asymmetry’s practical bite, and it is why the 2017 act’s durability profile looks the way it does. A future Congress that wants to preserve the individual provisions must pass a new law under whatever budget constraints then apply, which means the 2017 act’s most visible benefits to individual filers are durable only if a future coalition pays for their extension.

The comparison also corrects a common misreading of what durability means. Durability is not popularity and it is not virtue; it is the resistance of a provision to reversal, and resistance comes from different sources in the two statutes. In 1986, resistance came from the breadth of the coalition: with 97 senators on record, reversal meant confronting a near-consensus. In 2017, resistance for the corporate rate comes from the legal fact of permanence plus the political cost of raising a rate, while the individual provisions have almost no resistance at all beyond the hope of extension. The honest durability scorecard therefore reads: 1986 changed the base in a way that stuck and the rates in a way that did not; 2017 changed the corporate rate in a way that has stuck so far and the individual rates in a way designed not to. Neither statute achieved durable change across the board, and the reasons trace back to the first three axes: fiscal design determined what had to be traded, procedure determined what could be permanent, and distributional design determined who would defend the result.

Two Theories of Reform: Base First Versus Rate First

Beneath the four axes lies a disagreement about what tax reform is, and the two statutes embody opposite answers. The 1986 theory held that the base comes first: broaden the definition of taxable income by repealing preferences, and the broader base will finance lower rates without losing revenue or shifting the burden. The slogan of the era, broad base and low rates, was not marketing but mechanics, because the revenue-neutrality constraint made the slogan a theorem: with revenue fixed, the only way to cut rates was to broaden the base, so base broadening was not a preference but a prerequisite. The 1986 act’s intellectual coherence came from this tight coupling. Every provision could be justified as serving the single goal of a cleaner code taxed at lower rates, and the JCT’s balanced score was the proof that the coupling held.

The 2017 theory held that the rate comes first: cut the corporate rate to make the United States competitive for investment, cut individual rates to raise after-tax incomes, and let the deficit absorb what the offsets do not cover. Base broadening appeared in the 2017 act, through the SALT cap and the international guardrails, but as a secondary instrument for staying inside the $1.5 trillion instruction, not as the engine of the reform. The intellectual coherence of the 2017 act came from a different coupling: the rate cuts were the goal, the reconciliation instruction was the financing, and the sunsets were the price of the financing. Where 1986 asked what base would support the desired rates, 2017 asked what rates the permitted deficit would support.

The base-first theory has a durability advantage that follows from its logic. A broadened base is a structural fact about the code: once the investment tax credit is gone, investment decisions stop depending on it, and restoring it would require Congress to recreate a preference whose constituency has dispersed. The 1986 base broadeners thus benefited from a kind of political entropy: repeal is easier to sustain than to reverse because the beneficiaries of repeal, all taxpayers enjoying lower rates, are diffuse, while the beneficiaries of restoration would have to organize from scratch. The rate-first theory has the opposite entropy. A rate cut’s beneficiaries are concentrated and organized, which makes the cut politically durable in the short run, but the cut is financed by borrowing, which leaves the fiscal pressure in place for the next Congress to relieve by raising the rate. The 1990 and 1993 repricings of the 1986 rates illustrate the point from the other side: even a base-first reform could not protect its rates, because rates are the most reversible instrument in the code.

The two theories also imply different relationships between the tax code and economic decision-making, which is worth stating because it explains why economists divided over the two bills. The base-first theory treats the code’s complexity as the central economic cost: preferences distort investment toward tax-favored activities, so repealing them improves the allocation of capital even before rates fall. The 1986 act’s repeal of the investment tax credit and the passive loss limits were defended on exactly these grounds, as removing distortions that had steered capital into tax shelters. The rate-first theory treats the level of rates as the central economic cost: high marginal rates discourage work and investment at the margin, so cutting them improves incentives even if the base stays narrow. The 2017 act’s corporate rate cut was defended on these grounds, as improving the incentive to invest in the United States regardless of the preference structure. Both mechanisms are real, and the two bills can be read as wagers on which distortion mattered more. The 1986 wager was that base complexity was the binding constraint on efficiency; the 2017 wager was that the corporate rate was. The verdict of this article does not adjudicate the efficiency debate, which belongs to economics, not to legislative mechanics; it adjudicates only which wager produced the more durable legislative achievement.

Neither theory is refuted by the other’s existence; they optimize for different objectives under different constraints. The base-first theory optimizes for structural durability and is the right model when the goal is a cleaner code that outlasts the enacting coalition. The rate-first theory optimizes for immediate rate relief and is the available model when no majority will accept the neutrality constraints. The comparison’s verdict prefers the base-first theory for future reform precisely because future reform, if it is to deserve the name, should aim at structure, not at a scheduled cliff. But the verdict is conditional in the way the article has been explicit about: the base-first theory is available only at the neutrality price, and a Congress unwilling to pay it will find the rate-first theory waiting, with its characteristic costs attached.

Why the Votes Looked So Different

The vote counts are the most quoted fact about each statute and the least explained, so they deserve a synthesis that ties the four axes together. The 1986 conference report passed the House 292 to 136 and the Senate 74 to 23, with majorities of both parties in favor. The 2017 conference report passed the Senate 51 to 48 and the House 224 to 201, with the minority unanimously opposed. The explanation is not that one era was more collegial than the other; it is that each statute’s fiscal design, distributional design, and procedure jointly determined the coalition it needed and the coalition it got.

The gap between the 1986 Senate’s 97 to 3 pre-conference vote and its 74 to 23 conference-report vote is itself instructive and often misread. The 97 to 3 vote measured the Senate’s enthusiasm for its own bill, written in the Finance Committee with heavy minority participation and full of provisions senators had personally negotiated. The 74 to 23 vote measured the Senate’s verdict on the conference compromise, which had moved rates, adjusted the capital gains treatment, and settled dozens of smaller issues in ways that disappointed senators on both sides. That 23 senators defected from the conference report is not evidence of weakness but of the conference’s reality: a genuine compromise disappoints its authors, and the disappointed authors vote no. The important fact is that majorities of both parties still voted yes, because the underlying bargain, lower rates for a broader base under the two neutrality constraints, survived the conference. A reader who cites the 97 to 3 vote as the 1986 act’s final margin is making the exact error the fact-check corrects; a reader who understands why the margin narrowed from 97 to 3 to 74 to 23 understands how regular order actually works.

Start with what the 1986 majority needed. Regular order gave the minority the power to slow or block the bill, so passage required minority votes. Revenue neutrality gave the majority a reason to seek them: a bill that must be paid for inside its own text needs a broad market of preferences to trade, and the minority controlled preferences the majority wanted repealed. Distributional neutrality gave both sides a reason to stay at the table: with the burden distribution held fixed by design, neither party’s members had to explain to their voters why the bill had picked their side to pay. Each constraint thus did coalition work. Revenue neutrality forced trades, distributional neutrality prevented defections, and regular order made both constraints enforceable by giving the minority a veto over any bill that violated them. The supermajority was the output of that machine, not a tribute to an unusually cooperative Congress.

Consider what the 2017 majority needed, which was nothing from the minority. Reconciliation’s 51-vote threshold removed the minority’s leverage, so the majority did not have to buy minority votes. The reconciliation instruction’s $1.5 trillion allowance removed the need to trade: rate cuts could be financed by the permitted deficit increase instead of by repealing preferences the minority might have defended. And the absence of a distributional constraint removed the need to neutralize the winners-and-losers fight: the majority could accept a distributional profile the minority would attack, because the minority’s attack could not stop the bill. Each freedom thus did its own coalition work in reverse. The party-line vote was the output of a machine designed to need no minority at all, and the minority, having been given no reason to participate and no power to obstruct, did what minorities do under those conditions: it voted no and prepared its repeal arguments for the next election.

The asymmetry in ownership followed. The 1986 act belonged to both parties, which meant neither party could easily campaign against it and both parties had reasons to defend its base-broadening achievements even while later raising its rates. The 2017 act belonged to one party, which meant the other party could campaign against it freely and had no stake in defending its individual provisions when the sunsets approached. This is the coalition-building half of the verdict the article will reach: a reform that wants durable structural change needs a coalition broader than one party, and the constraints that build such a coalition are revenue neutrality and distributional neutrality, because those are the constraints that force each side to name what it will give up.

The partisan configuration of each Congress sharpened these dynamics without creating them. In 1986, divided government, a Democratic House, a Republican Senate, and a Republican president, meant no single party could legislate alone even under the most majoritarian procedure available; regular order merely formalized a necessity the election results had already imposed. President Reagan’s support gave Republicans cover for the preference repeals, while Democratic control of the House gave Democrats ownership of the final product. In 2017, unified Republican control of the House, the Senate, and the presidency meant the majority faced no constitutional obstacle to partisan legislation; reconciliation merely formalized an opportunity the election results had already created. The procedures thus amplified the underlying partisan arithmetic instead of inventing it. But amplification is not determination: divided governments have passed party-line bills through reconciliation, and unified governments have pursued bipartisan regular-order bargains. The 1986 and 2017 majorities each chose the procedure that served their substantive goals, and the goals, durable structural reform versus rapid rate relief, are what the four axes capture.

This distinction matters because it answers the most common misreading of the vote counts. The misreading treats 97 to 3 and 51 to 48 as measures of congressional collegiality, as though the 1986 Congress was simply friendlier than the 2017 Congress. The vote counts measure something more mechanical: the size of the coalition each bill’s design required. The 1986 bill required a large coalition because its procedure gave the minority leverage and its constraints gave the minority reasons to bargain; the large coalition then appeared in the vote. The 2017 bill required no coalition beyond the majority because its procedure removed the minority’s leverage and its lack of constraints removed the minority’s reasons to bargain; the party-line vote then appeared. To ask why the 2017 Congress was less bipartisan than the 1986 Congress is to ask the wrong question. The right question is why each Congress chose the procedure and constraints it chose, and the answer in both cases is that the majority selected the legislative technology that would produce the substantive result it wanted.

The 1986 Bargain in Detail

The 1986 act did not emerge from a single negotiation. It survived a sequence of near deaths, each of which illustrates how the neutrality constraints operated as the bargaining rules. The Treasury Department released its initial reform proposals in late 1984, and the administration’s revised plan followed in 1985. Senator Bill Bradley, a Democrat from New Jersey, had already been developing base broadening legislation with Representative Richard Gephardt, which gave the effort a bipartisan intellectual foundation before the White House plan arrived. That bipartisan parentage mattered later, when the bill needed defenders in both parties.

The House phase tested whether a Democratic committee chairman would carry a Republican president’s bill. Representative Dan Rostenkowski, chairman of the Ways and Means Committee, moved H.R. 3838 through markup in the fall of 1985, holding together a coalition that included liberals who wanted a fairer code and conservatives who wanted lower rates. The House passed the bill by voice vote on December 17, 1985. A voice vote in this context is itself evidence of the coalition’s breadth: no member demanded a recorded vote because no faction wanted to be recorded against a bill its own party’s chairman had built.

The Senate phase nearly killed the bill. The Senate Finance Committee under Chairman Bob Packwood deadlocked over the distribution of the base broadening, and the legislation stalled in the spring of 1986. Packwood revived it by offering deeper rate cuts financed by deeper base broadening, a move that worked precisely because the revenue neutral framework made the trade legible. Lower rates were available to anyone willing to name the preferences that would pay for them, and once Packwood named a package, the committee moved. The Senate passed the bill as amended 97 to 3 on June 24, 1986, a margin that reflected how the neutrality constraints had converted opponents into negotiators.

The conference committee then reconciled the House and Senate versions through the summer and early fall. Conference bargaining under the neutrality rules was an exchange of specific provisions: the House’s preferences against the Senate’s, each chamber’s rate ambitions against the other’s base broadening. Because neither chamber could lose revenue or shift the distribution, every concession had to be matched, and the matching is what produced the final package’s coherence. The conference report passed the House 292 to 136 on September 25 and the Senate 74 to 23 on September 27, and President Reagan signed the bill on October 22, 1986. The full sequence, from Treasury proposal to signature, took nearly two years, which is the normal pace of regular order when the minority must be accommodated at every stage.

The lesson of the sequence is that the 1986 constraints did not merely limit the bill. They organized the bargaining that produced it. Revenue neutrality told every participant where the money for rate cuts would come from. Distributional neutrality told every participant that the bill would not be used against their constituents. Regular order gave the minority the leverage to enforce both rules. Remove any one of the three and the bargain collapses, which is exactly what the 2017 procedure demonstrates in reverse.

The Two Coalitions Up Close

The contrast between the coalitions is best seen in the people who built them, reported here by name and office as matters of record. The 1986 coalition began with President Ronald Reagan, who supplied the political demand for lower rates and the willingness to sign a bill that raised corporate revenue to pay for them. Treasury Secretary James Baker supplied the institutional design, translating the president’s rate cutting ambition into a revenue neutral framework the committees could actually write. Senator Bill Bradley, Democrat of New Jersey, supplied the bipartisan intellectual foundation, having developed base broadening proposals with Representative Richard Gephardt before the administration’s plan arrived. Representative Dan Rostenkowski, Democrat of Illinois and chairman of the House Ways and Means Committee, carried the bill through the House, holding together liberals who wanted a fairer code and conservatives who wanted lower rates. Senator Bob Packwood, Republican of Oregon and chairman of the Senate Finance Committee, revived the bill after its near death in committee and steered it to the 97 to 3 Senate passage. Each of these participants surrendered something: Reagan accepted the corporate revenue increase, Baker accepted the slow grind of regular order, Bradley accepted rates lower than many Democrats wanted, Rostenkowski accepted the elimination of preferences his party’s constituencies valued, and Packwood accepted a bill far more ambitious than the one he had started with.

The 2017 coalition was narrower by design and by procedure. President Donald Trump supplied the political demand for a large tax cut and signed the bill on December 22, 2017. House Speaker Paul Ryan and Senate Majority Leader Mitch McConnell managed the floor strategy that made reconciliation work, keeping their conferences unified through the compressed schedule. Representative Kevin Brady, Republican of Texas and chairman of the House Ways and Means Committee, wrote the House bill in November. Senator Orrin Hatch, Republican of Utah and chairman of the Senate Finance Committee, wrote the Senate version. On the other side of the procedural fights stood Senator Bernie Sanders, Independent of Vermont, whose Byrd rule point of order struck the bill’s short title, and Senator Mike Enzi, Republican of Wyoming, whose waiver motion failed 51 to 48. No Democrat in either chamber voted for the conference report, which means the 2017 coalition, unlike the 1986 one, contained no one who had to be persuaded across party lines and no one who surrendered a preference to get there.

The comparison of the rosters is the comparison of the designs in human form. The 1986 roster includes committee chairmen from both parties because regular order required both parties. The 2017 roster is a single party’s leadership because reconciliation required only a majority. The 1986 participants each paid for their rate cuts with preferences their own supporters valued, because neutrality demanded it. The 2017 participants paid with a deficit increase, because the instruction permitted it. Coalitions are not accidents of personality. They are the predictable output of the rules, and the rules for the two bills were opposites.

The Afterlife: What Later Congresses Did With Each Statute

Durability is proven after passage, so the afterlife of each statute belongs in the comparison. The 1986 act’s afterlife confirmed the axis-four prediction: the base endured while the rates were repriced. The Omnibus Budget Reconciliation Act of 1990 replaced the 5 percent surcharge with a 31 percent statutory top rate, and the Omnibus Budget Reconciliation Act of 1993 added the 36 and 39.6 percent brackets, per CRS RL34498. Both were deficit-reduction bills that needed revenue, and both found it easiest to raise the top rates on the broad 1986 base instead of narrowing the base again. Later Congresses tinkered at the edges, restoring a capital gains differential and adjusting the alternative minimum tax, but the core 1986 architecture, a broad definition of taxable income with limited preferences, survived every subsequent rewrite. The base broadeners had done their work so thoroughly that even Congresses hostile to the 1986 rates kept the 1986 base.

The 2017 act’s afterlife is still being written, and its structure tells us what to watch. The corporate rate cut to 21 percent has survived multiple Congresses and shows no sign of reversal, which is consistent with the durability of permanent, deficit-financed rate cuts whose beneficiaries are organized. The individual provisions face their scheduled expiration after December 31, 2025, and the extension debate will replay the original fiscal-design choice: extending them requires either new offsets or a new deficit allowance, scored by the JCT under whatever budget rules then apply. The chained-CPI indexing, being permanent, will keep operating silently in the background, gradually raising real burdens relative to the old inflation measure regardless of what happens to the sunsets. And the international provisions will step up after 2025, with GILTI, FDII, and BEAT rates rising as scheduled, which means the 2017 act’s base-related architecture tightens automatically while its individual relief expires automatically. The statute was designed as a machine with moving parts, and the parts are moving on schedule.

The afterlives also illustrate the coalition point. The 1986 base survived because both parties owned it: Democrats who had voted for the preference repeals could not easily campaign for their restoration, and Republicans who had voted for the rate cuts defended the base that financed them. The 2017 individual provisions have no such cross-party ownership, which means their extension depends entirely on the majority of the moment. A future Congress controlled by the enacting party may extend them; a future Congress controlled by the other party may let them expire and blame the original sunsets. Either way, the provisions’ fate will be decided by partisan control, not by structural consensus, which is the predictable afterlife of a party-line bill. The contrast with the 1986 base, which survived changes of party control without serious challenge, is the durability axis rendered as history.

There is a final afterlife lesson about the interaction of the axes. The 1986 act’s revenue neutrality meant its afterlife imposed no fiscal hangover: later Congresses inherited a code that raised roughly the revenue it was designed to raise, and their rate increases were choices, not necessities. The 2017 act’s deficit financing means its afterlife carries a fiscal overhang: the $1.46 trillion conventional score was the down payment, and any extension of the individual provisions adds to it. Future Congresses will therefore face the 2017 act not as a settled structure but as an open question, how much of the temporary law to make permanent and how to pay for it. That open question is the sunset asymmetry’s long-run cost, and it will keep the 2017 act in active political contention for as long as the expirations remain unresolved.

Sunsets as Strategy

The 2017 sunsets deserve analysis as a deliberate legislative strategy instead of a mere accounting necessity, because they do political work that extends far beyond the budget window. A sunset converts a one time enactment into a recurring event. When the individual provisions expire after 2025, the expiration does not quietly restore prior law. It creates a scheduled crisis in which taxes rise for millions of households unless Congress acts, which gives the enacting party a permanent talking point and the opposing party a permanent dilemma. Voting against extension becomes a vote for a tax increase, which is a difficult vote for any legislator regardless of party. The sunset is therefore not only a way to fit the bill under the Byrd rule. It is a way to project the majority’s power into future Congresses that the majority may not control.

This strategy has a distinguished and controversial pedigree in tax legislation. The 2001 and 2003 tax cuts used the same device, with provisions scheduled to expire after 2010, and the resulting cliff dominated tax politics for a decade. The 2017 drafters knew this history and used the device with open eyes, allocating permanence to the corporate rate, whose beneficiaries can defend it without a scheduled crisis, and expiration to the individual provisions, whose beneficiaries need the crisis to force action. The asymmetry is the strategy: protect what can protect itself, and schedule a fight over the rest.

The 1986 act used the opposite approach to permanence. Its base provisions were enacted without expiration dates, which meant their survival depended on no future vote. The passive loss rules, the repealed investment credit, and the other base broadeners persisted because repealing a repeal requires affirmative legislation, and no coalition ever assembled to do it. The 1986 rates, enacted as permanent law in form, proved temporary in fact, because changing a rate requires only a new number and every budget bill needs revenue. The lesson is that statutory permanence and political permanence are different things. The 1986 base was politically permanent without needing the word. The 2017 individual provisions are politically contingent despite being called tax cuts, because their continuation requires votes that have not yet been cast.

A future reformer should weigh these two theories of permanence carefully. The 1986 theory says: eliminate preferences outright, create no scheduled events, and let the new status quo defend itself through the reorganization of interests around it. The 2017 theory says: enact what you can permanently, sunset the rest, and use the scheduled crises to keep your coalition mobilized. The first theory produced base provisions that survived decades. The second theory produced a decade of scheduled cliff politics in its earlier incarnation. Which theory serves a given reform depends on whether the reformer values structural quiet or political leverage, and the two statutes offer the clearest available evidence for the tradeoff.

The Complication: Was the 2017 Act Just 1986 With Different Numbers?

A persistent claim holds that the 2017 act was simply the 1986 act with different numbers: both cut the top corporate rate, both lowered individual rates, both were sold as pro-growth reform, so the differences are cosmetic. The claim is tempting because the rate-cut headlines rhyme, and it is wrong in every dimension the four axes measure. The two statutes used opposite fiscal designs: 1986 was revenue neutral by construction, with about $122 billion of individual revenue loss matched against about $120 billion of corporate revenue gain, while 2017 was enacted under an instruction permitting up to $1.5 trillion in added deficits and scored at about $1.46 trillion in added deficits on a conventional basis. They used opposite procedures: 1986 moved under regular order through both committees and a conference with minority support at every stage, while 2017 moved through reconciliation with a 51-vote threshold, no minority support, a struck short title, and a failed waiver vote. They used opposite durability strategies: 1986 wrote permanent base broadening and permanent rates that a later Congress repriced, while 2017 wrote a permanent corporate rate alongside individual provisions designed to expire.

The only substantial similarity is that both cut the top corporate rate, and even that similarity dissolves on inspection. The 1986 corporate cut, from 46 to 34 percent, was largely paid for with corporate base broadeners, including the repeal of the investment tax credit. The 2017 corporate cut, from 35 to 21 percent, was not paired with equivalent offsets; the JCT scored the rate reduction alone at a loss of about $1,348.5 billion over the window, with the international redesign’s revenue gains, $338.8 billion from the section 965 transition tax, $112.4 billion from GILTI, and $149.6 billion from BEAT, offsetting only a fraction of it. One was a financed rate cut inside a broader base; the other was a deficit-financed rate cut alongside a restructured international system. To call them the same reform with different numbers is to compare the price tags while ignoring what was purchased and how it was paid for.

The claim persists because it serves both sides’ rhetoric. Supporters of the 2017 act invoked 1986 to borrow its bipartisan prestige; critics of the 2017 act invoked 1986 to highlight what bipartisanship used to require. Both invocations treat 1986 as a brand, not a mechanism. The structural comparison shows the brand is not transferable: the 1986 result came from the 1986 constraints, and a statute that declines the constraints cannot claim the result. That is the point the verdict makes explicit.

The numbers sharpen the refutation. Compare the two statutes’ signature achievements as the JCT scored them. The 1986 act’s individual rate cuts lost about $122 billion over five years and were fully offset within the bill; the 2017 act’s individual rate cuts lost about $1,214.2 billion over ten years and were offset only in small part. The 1986 corporate rate cut was more than paid for by corporate base broadeners, producing a net corporate revenue gain of about $120 billion over five years; the 2017 corporate rate cut lost about $1,348.5 billion over ten years against international offsets totaling a few hundred billion. These are not different numbers for the same reform. They are opposite fiscal architectures: one in which rate cuts were purchased with base broadening, and one in which rate cuts were purchased with borrowing. A reader who grasps that distinction will never again be tempted by the different-numbers claim, because the claim requires ignoring the only part of each bill’s score that explains its shape.

There is a narrower version of the claim worth addressing separately: that the 2017 act’s international provisions were the true heir of the 1986 base-broadening spirit, since both restructured how business income is taxed. The international redesign was indeed the 2017 act’s most structural achievement, and GILTI, FDII, and BEAT did broaden the international base in the sense of taxing income that had previously escaped. But the comparison still fails on fiscal design: the 1986 base broadening financed rate cuts within a neutral total, while the 2017 international provisions financed only a fraction of a much larger rate cut within a $1.46 trillion deficit increase. Base broadening that pays for a quarter of the rate cuts is a different fiscal animal from base broadening that pays for all of them. The 2017 international title deserves study as international tax policy; it does not deserve the 1986 mantle.

The Corporate Rate as Common Ground

The one substantial similarity between the statutes deserves its own examination, because it is the similarity the complication section sets aside and it repays closer attention. Both acts cut the top corporate rate: from 46 percent to 34 percent in 1986, and from 35 percent to 21 percent in 2017. In both cases the corporate cut was the provision with the broadest elite support, drawing on arguments about competitiveness, investment, and the efficiency costs of a high statutory rate on a narrow base. In both cases the cut was paired with base changes meant to offset part of its cost. The resemblance is real, and a reader who notices it is not imagining things.

But the two cuts were built for different purposes inside different designs, and the differences matter more than the resemblance. The 1986 corporate cut was the individual side’s mirror: the corporate base was broadened by about 120 billion dollars over five years to help pay for the individual rate cuts, and the corporate rate fell as compensation for the broader base. The cut was embedded in a neutral package, which meant its size was limited by the offsets available. The 2017 corporate cut was the package’s centerpiece: it lost about 1,348.5 billion dollars over ten years, the largest single cost in the bill, and it was made permanent while everything else sunset. Its size was limited only by the reconciliation cap, not by offsets, which is why it could be so much larger relative to the base provisions around it.

The durability of the two cuts also diverged, for reasons the fourth axis explains. The 1986 corporate rate of 34 percent was itself later changed by subsequent legislation, though the corporate base broadening that accompanied it survived. The 2017 corporate rate of 21 percent was designed as the permanent achievement of the bill, defended by the constituency it created. One cut was a component of a neutral trade. The other was the prize the deficit allowance purchased. To call them the same policy is to confuse the direction of the cut with the design of the bill, and the comparison exists to keep those two things distinct.

There is a final irony worth noting. The corporate rate is the ground on which the two parties have most often agreed, in 1986 with bipartisan votes and in 2017 with business community support that crossed party lines in the private sector even as the congressional vote did not. If a future reform seeks the widest possible coalition, the corporate rate is the natural starting point, because it is the provision both statutes prove can command support. But the history also shows that agreement on the rate does not produce agreement on the design: 1986 paired its cut with neutrality and bipartisanship, while 2017 paired its cut with a deficit allowance and party line votes. The rate is common ground. Everything built on top of it is a choice.

Answering the Strongest Objections

A verdict this direct invites objections, and the strongest ones deserve answers, not dismissal. Three objections recur whenever the 1986 model is held up against the 2017 model, and each tests a different part of the argument.

The first objection holds that the 1986 coalition was a one-time product of divided government that cannot be repeated. In 1986 a Democratic House, a Republican Senate, and a Republican president each needed the others, so bipartisanship was forced by the partisan configuration, not chosen through the constraints. In 2017 unified Republican control made bipartisanship unnecessary, so any unified government would have done the same. The objection has force as description but fails as refutation, because it concedes the mechanism while disputing its repeatability. If divided government forced the constraints, then the constraints still did the work: revenue and distributional neutrality were what the divided parties negotiated through, and the resulting coalition and durability are still attributable to them. Moreover, unified governments have repeatedly chosen regular order and bipartisan tax bargains when they wanted durable results; the choice of reconciliation in 2017 was a choice, not a law of nature. The objection therefore relocates the decision without changing its terms: even if divided government made the 1986 choice easier, the choice remains the one a future Congress must make.

The second objection holds that reconciliation has become the normal vehicle for major tax legislation, so the 1986 model is unrepeatable regardless of its merits. There is truth in the premise: the filibuster’s effective 60-vote threshold has made regular-order tax reform extraordinarily difficult, and both parties have used reconciliation for their signature tax bills. But the objection proves too much, because it treats a procedural equilibrium as permanent. Procedural equilibria change when their costs become visible, and the 2017 act made reconciliation’s costs unusually visible: the sunsets, the struck title, the party-line ownership, and the scheduled cliff that forces every future Congress to relitigate the individual provisions. A future majority that wants durable structural change may rationally conclude that the price of regular order, time and concessions, is lower than the price of reconciliation, impermanence and partisan ownership. The 1986 model is unrepeatable only so long as majorities prefer speed to durability; the moment a majority’s preference reverses, the model’s requirements, the two neutrality constraints, are still on the shelf.

The third objection holds that distributional neutrality is itself a contestable substantive choice disguised as a procedural constraint, because holding the distribution fixed entrenches whatever distribution prior law produced. This is the most serious objection, and the article’s answer is to embrace the premise while limiting its scope. Distributional neutrality is indeed a substantive choice: it says the tax reform will not be the vehicle for redistribution. But it is a choice about the bill, not about the tax system; a Congress that wants a different distribution can legislate it separately, through provisions whose distributional effects are debated on their own merits, not smuggled inside a reform bill. The 1986 negotiators understood this separation: distributional neutrality kept the reform coalition together precisely because it moved the redistribution fight to other legislation. A future reform can pair distributional neutrality in the reform bill with explicit distributional legislation alongside it, achieving both structural reform and a chosen distribution without asking one bill to do both jobs. The objection thus identifies a real cost of the constraint and simultaneously points to the remedy.

A fourth, quieter objection holds that the verdict overstates the 1986 achievement because its rates did not survive either. If both statutes’ rate structures proved impermanent, the objection runs, then the 1986 model’s durability advantage reduces to its base broadening, which is a thinner claim. The answer is that the thinner claim is still decisive, because base broadening was the point. The 1986 act set out to rebuild the definition of taxable income, and that rebuild survived: the repealed investment tax credit, the section 469 passive loss limits, the at-risk rules, and the repealed consumer interest deduction all remain features of the code. The rates were always the adjustable parameter, repriced by the 1990 and 1993 acts without disturbing the base. A reform that durably improves the base while leaving rates to later bargaining has succeeded on its own terms; a bill that cuts rates durably while leaving the base narrow has succeeded on different terms. The verdict prefers the former because structure outlasts pricing, and the afterlife of both statutes confirms the preference.

The Verdict: Which Model Serves a Future Reform

The comparison resolves into a verdict with a named deciding factor. On structural durability and coalition-building, the 1986 model is the better model for a future reform, and the deciding factor is the constraint: revenue neutrality paired with distributional neutrality. The 1986 model produces more durable structural change, a broader and more defensible base, and a coalition that survives the next election, at the price of requiring both constraints. No modern majority has been willing to accept that price, which is why no Congress since 1986 has repeated the 1986 result. The honest answer names the trade directly.

The honesty of the answer matters because the tax debate is saturated with costless superlatives. Every major tax bill is sold as reform, every rate cut as growth, every coalition as broad. The four-axis comparison cuts through the salesmanship by asking what each bill’s design required and what it produced. The 1986 act required the neutrality price and produced a durable broader base with bipartisan ownership. The 2017 act declined the price and produced a permanent corporate rate cut with temporary individual relief and single-party ownership. Neither outcome was mysterious given the inputs, and a future Congress that wants the 1986 outcome must supply the 1986 inputs. The verdict is therefore less a ranking of past bills than a recipe for future ones: if durable structural change is the goal, the constraints are the method, and the method has a known price.

The case for the 1986 model rests on the durability record. Its base broadening survived: the repealed investment tax credit, the passive loss limits codified at section 469, the at-risk rules, and the repeal of the consumer interest deduction all remain features of the code decades later. Its coalition survived too, in the sense that both parties owned the structure and neither campaigned to restore the repealed preferences. What did not survive was the rate structure, and that failure is instructive, not disqualifying: the 1990 and 1993 budget acts raised the top rates while keeping the broadened base, which means the 1986 reform delivered exactly what its design promised, a better base, even as later Congresses repriced it. A future reform that copies the 1986 model should therefore expect the same pattern: durable improvement in the structure of the code, with rates remaining the perennial subject of later bargaining.

The case against copying the 2017 model as a reform template is not that the 2017 act failed on its own terms; it is that its terms were different from reform’s. The 2017 act delivered a large permanent corporate rate cut and temporary individual relief through a procedure that required no minority buy-in and no offsets. That is a legitimate use of reconciliation, but it is tax cutting, not tax reform in the 1986 sense, because it left the preference structure largely intact and left the individual provisions to expire by design. Its durability profile reflects that choice: the corporate rate persists because it was written permanently, while the individual provisions persist only if a future Congress extends them. A future reform that wants structural change, not a scheduled cliff, should not copy the sunset strategy, because sunsets are what a majority writes when it cannot assemble the coalition for permanence.

The neutrality price: bipartisan tax reform is available only at the price of revenue and distributional neutrality, because those constraints are what force each side to name what it will give up, and every tax bill that has declined to pay that price has passed on party lines.

That claim is the article’s defended verdict, and it carries two implications a future Congress should face squarely. First, the price is real and must be budgeted politically, not just fiscally: accepting revenue neutrality means telling rate-cut supporters which preferences will die, and accepting distributional neutrality means telling both parties’ constituencies that the bill will not redistribute the burden. Second, declining the price is also a choice with known consequences: a bill that will not pay for its rate cuts and will not hold distribution fixed can still pass, but it will pass on party lines, through reconciliation, with sunsets, and its achievements will belong to one party and expire on a schedule. Neither path is dishonest; only pretending that one path delivers the other’s results is dishonest. The 1986 model remains the better model for durable structural reform, and the constraint remains the deciding factor, because the constraint is what makes the coalition, and the coalition is what makes the change last.

What would following the 1986 model mean operationally for a future Congress? It would mean beginning with the two constraints stated as non-negotiable: the bill must be revenue neutral as scored by the JCT over the budget window, and it must be distributionally neutral as measured by the JCT’s distributional tables. Those constraints would then dictate the procedure, because only regular order gives the minority the leverage that makes the constraints enforceable and the coalition possible. The tax-writing committees would mark up under those rules, trading preferences for rates provision by provision, with the JCT updating the scores as the bill changed. The conference would reconcile the chambers’ versions under the same constraints. The final votes would be bipartisan not because bipartisanship was the goal but because the constraints made it the byproduct. This is not nostalgia; it is a legislative technology, and like any technology it works whenever its operating conditions are met.

The chief obstacle is the one the article has named throughout: no modern majority has been willing to pay the neutrality price. Accepting revenue neutrality means telling every rate-cut constituency which preferences will die to finance the cuts, and modern majorities prefer to finance cuts with deficits. Accepting distributional neutrality means telling both parties that the bill will not redistribute the burden, and modern majorities prefer bills that deliver visible gains to their own coalition. These preferences are understandable, and they explain the 2017 choice. But they are preferences, not necessities, and a future majority that ranks durable structural change above immediate partisan advantage can still choose differently. The 1986 precedent proves the technology works; it does not prove that any particular Congress will use it. The verdict is therefore addressed to a hypothetical future majority with its priorities ordered toward durability, and it tells that majority exactly what the price is and why the price is worth paying.

One boundary on this verdict needs stating plainly. This article takes no position on which distribution of the tax burden is preferable. Its judgment is restricted to structural durability and coalition-building: which procedures and constraints produce reforms that endure and command broad ownership. Readers who favor a more or less progressive distribution than either statute produced should treat the verdict as a claim about legislative mechanics, not about fairness, and should demand the same sourced estimates, JCT, CBO, and Tax Policy Center with their windows, from any future bill that asks for their support.

How to Study These Two Statutes

Readers who want to go deeper should study the statutes in an order that matches the comparison’s logic. Start with the 1986 act’s provisions and passage, because its mechanism, constraints producing coalition producing durability, is the baseline against which the 2017 act’s departures become visible. Then study the 2017 act’s provisions with the sunset structure in mind, provision by provision, asking of each one whether it is permanent or temporary and what budget rule forced the choice. Then return to the four axes and test yourself against the article’s one test: state the four differences, explain the two coalitions, say which statute changed the base and which changed the rates, and defend a verdict. The US legislation study guide offers a recommended sequence for working through the series’ tax titles in that order.

Two companion tools support that study path. The VaultBook legislation study notebook gives readers a structured place to record each statute’s provisions, votes, and scores side by side, which is the exercise that turns the four-axis table from a diagram into knowledge. The ReportMedic civics study tool helps readers drill the procedural vocabulary, reconciliation, the Byrd rule, regular order, and conference, that the comparison assumes. Used together, the article, the study guide, and the two tools carry a reader from headline familiarity to structural understanding, which is the level at which the next tax debate will actually be joined.

A worked self-test shows what structural understanding looks like in practice. Take the sunset asymmetry and explain it three ways: as fiscal design, the individual provisions expire because the $1.5 trillion instruction could not absorb them permanently; as procedure, they expire because the Byrd rule forbids deficit increases beyond the budget window; and as durability, they expire because the enacting coalition could not assemble the price of permanence. Then do the same for the 1986 base broadening: as fiscal design, the preferences were repealed because revenue neutrality required offsets; as procedure, the repeals survived because regular order gave both parties ownership; and as durability, the repeals stuck because the ratchet effect destroyed the restoration constituencies. A reader who can run both statutes through all three explanations without consulting the article has internalized the four axes, and that reader is ready for the next tax bill, whatever its slogans claim.

A Reader’s Checklist: The One Test

This article set a one test at the outset: a reader who finishes it should be able to state the four structural differences, explain why one statute produced a bipartisan supermajority and the other a party-line vote, say which durably changed the tax base and which changed the rates, and reach a defended verdict on the better model for a future reform. The checklist below restates the test as four questions with the answers the article defends, so readers can verify their own understanding before moving on.

First, the four differences. Fiscal design: the 1986 act was revenue neutral by construction, with about $122 billion of individual revenue loss matched against about $120 billion of corporate revenue gain, while the 2017 act was enacted under a reconciliation instruction permitting up to $1.5 trillion in added deficits and scored at about $1.46 trillion. Distributional design: the 1986 act treated distributional neutrality as a binding design constraint, while the 2017 act accepted no distributional constraint, producing the Tax Policy Center’s finding of a $51,000 average 2018 cut for the top 1 percent and a 2027 landscape in which that group captures 83 percent of the remaining benefit. Procedure: the 1986 act moved under regular order through both committees and a conference, while the 2017 act moved through reconciliation with a 51-vote threshold and a Byrd-rule point of order that struck its short title. Durability: the 1986 base broadening survived while its rates were repriced in 1990 and 1993, while the 2017 corporate rate cut is permanent and its individual provisions expire after 2025.

Second, the coalitions. The 1986 supermajority, 292 to 136 in the House and 74 to 23 in the Senate on the conference report, resulted from regular order’s minority leverage combined with the two neutrality constraints, which gave the minority reasons to bargain and the majority reasons to buy minority votes. The 2017 party-line vote, 51 to 48 in the Senate and 224 to 201 in the House, resulted from reconciliation’s 51-vote threshold combined with the absence of the constraints, which removed both the need and the reason for minority participation.

Third, base versus rates. The 1986 act durably changed the tax base: the repealed investment tax credit, the section 469 passive loss limits, the at-risk rules, and the repealed consumer interest deduction all survived. The 2017 act durably changed the corporate rate, cutting it from 35 to 21 percent permanently, while its individual rate changes expire after 2025. Neither statute durably changed everything; their durability profiles are opposites.

Fourth, the verdict. The 1986 model is the better model for a future reform that seeks durable structural change, and the deciding factor is the constraint: revenue and distributional neutrality are the price of bipartisan tax reform, because those constraints force each side to name what it will give up. A reader who can state all four answers, with the scores and votes attached, has passed the test and is ready for the next tax debate.

Frequently Asked Questions

Q: What is the difference between the Tax Reform Act and the Tax Cuts and Jobs Act?

The Tax Reform Act of 1986, Public Law 99-514, was revenue neutral by construction: it cut the top individual statutory rate from 50 to 28 percent, with a 5 percent surcharge producing an effective 33 percent top marginal rate, and the corporate rate from 46 to 34 percent, paying for the cuts by repealing the investment tax credit, taxing capital gains as ordinary income, and limiting passive losses. The JCT scored the individual side at a loss of about $122 billion against a corporate gain of about $120 billion, a net of roughly negative $2 billion. It moved under regular order and passed with bipartisan supermajorities. The Tax Cuts and Jobs Act, Public Law 115-97, cut the corporate rate from 35 to 21 percent permanently and individual rates only through 2025, was scored by the JCT at about $1.46 trillion in added deficits over fiscal years 2018 through 2027, and passed through reconciliation on party-line votes. The structural differences run across fiscal design, distributional design, procedure, and durability, not just the size of the rate cuts.

Q: Which was bigger, the Tax Reform Act or the Tax Cuts and Jobs Act?

Bigness depends on the measure, and the two statutes win on different ones. As a fiscal event, the 2017 act was bigger: the JCT scored it at about $1.46 trillion in added deficits over ten years, while the 1986 act was scored at a net of roughly negative $2 billion over five years, essentially revenue neutral. The 2017 corporate rate cut alone, from 35 to 21 percent, was scored at a loss of about $1,348.5 billion over the window. As a structural rewrite, the 1986 act was bigger: it repealed the investment tax credit, ended the capital gains exclusion, imposed passive loss limits, and rebuilt the individual rate schedule around two statutory rates, while the 2017 act left most domestic preferences in place and concentrated its structural change in the international system. Measured by deficit impact, 2017 dwarfs 1986; measured by how much of the code’s architecture changed, 1986 dwarfs 2017.

Q: Was the Tax Cuts and Jobs Act bipartisan like the Tax Reform Act?

No. The 1986 act’s conference report passed the House 292 to 136 on September 25, 1986, and the Senate 74 to 23 on September 27, 1986, with majorities of both parties in favor, after a 97 to 3 pre-conference Senate vote on June 24. The 2017 act passed the Senate 51 to 48 and the House 224 to 201 on December 20, 2017, with every member of the minority party voting no in both chambers. The difference traces to procedure. The 1986 act moved under regular order, which gave the minority leverage at every stage and forced the majority to build a cross-party coalition. The 2017 act moved through reconciliation, whose 51-vote threshold let the majority pass the bill alone, so it did. Bipartisanship in 1986 was not a mood; it was the output of a procedure that required minority votes.

Q: Why was the Tax Reform Act revenue neutral but the Tax Cuts and Jobs Act not?

Because the two Congresses chose opposite budget procedures. The 1986 act was negotiated under a political commitment to revenue neutrality that the JCT’s scores enforced: about $122 billion of individual revenue loss matched against about $120 billion of corporate revenue gain, for a net of roughly negative $2 billion. Every rate cut had to be financed by a repealed preference inside the same bill. The 2017 act was written to reconciliation instructions in H. Con. Res. 71, section 2001(a), which permitted the deficit to increase by not more than $1.5 trillion for fiscal years 2018 through 2027. The JCT scored the enacted bill at about $1.46 trillion in added deficits on a conventional basis, just inside the allowance. Revenue neutrality was a choice the 1986 majority made and the 2017 majority declined, and each choice was implemented through the budget procedure that made it possible.

Q: Did the Tax Reform Act or the Tax Cuts and Jobs Act broaden the base more?

The 1986 act broadened the base far more. It repealed the investment tax credit, repealed the 60 percent exclusion for long-term capital gains, imposed strict passive activity loss limits codified at section 469, which IRS Publication 925 confirms remain current law, curtailed at-risk deductions, repealed the consumer interest deduction, and expanded the alternative minimum tax. The 2017 act’s base broadening was narrower: the $10,000 cap on the state and local tax deduction, scored at a gain of about $670.6 billion over ten years, the suspension of the personal exemption to zero through 2025, the repeal of the domestic production activities deduction, and the international redesign, whose gains from the section 965 transition tax, GILTI, and BEAT partly offset the corporate rate cut. The 2017 act kept the great majority of domestic preferences intact, which is what its fiscal design allowed.

Q: Which lasted longer, the Tax Reform Act or the Tax Cuts and Jobs Act rates?

It depends on which rates. The 1986 individual rate structure lasted about four years: the 28 percent statutory rate with its 5 percent surcharge gave way to a 31 percent statutory rate under the 1990 budget act, and the 1993 budget act added 36 percent and 39.6 percent brackets. The 2017 corporate rate cut to 21 percent is permanent and has already outlasted the 1986 individual rate structure, persisting until Congress affirmatively raises it. The 2017 individual rates, by contrast, are scheduled to expire after 2025, giving them an eight-year scheduled life from 2018 through 2025. The honest scorecard is therefore split: 2017 wins on corporate rate durability, 1986’s base changes outlasted both statutes’ individual rate changes, and the 2017 individual rates survive only if a future Congress extends them.

Q: Did the Tax Cuts and Jobs Act follow the Tax Reform Act model?

No. The claim that the 2017 act was the 1986 act with different numbers fails on every structural axis. Fiscal design was opposite: revenue neutral by construction in 1986, with a net score of roughly negative $2 billion, versus a permitted $1.5 trillion deficit increase in 2017, scored at about $1.46 trillion. Procedure was opposite: regular order with minority support in 1986 versus reconciliation on party lines in 2017, complete with a Byrd-rule point of order that struck the 2017 bill’s short title. Durability strategy was opposite: permanent base broadening with repriced rates in 1986 versus a permanent corporate rate with expiring individual provisions in 2017. The only substantial similarity is that both cut the top corporate rate, and even there the 1986 cut was largely paid for while the 2017 cut was deficit-financed.

Q: Which should a student study first, the Tax Reform Act or the Tax Cuts and Jobs Act?

Study the 1986 act first. Its mechanism, revenue and distributional neutrality as constraints producing a bipartisan coalition producing durable base broadening, is the baseline against which the 2017 act’s departures become legible. Learn the 1986 provisions, the JCT’s roughly neutral score, the regular-order path from the December 1985 House vote through the September 1986 conference report, and the 1990 and 1993 repricing of its rates. Then study the 2017 act provision by provision, asking of each whether it is permanent or temporary and which budget rule forced the choice. Finish by testing yourself against the four axes: state the fiscal, distributional, procedural, and durability differences, explain the two coalitions, and defend a verdict on the better model for a future reform.

Q: What role did the Byrd rule play in the Tax Cuts and Jobs Act?

The Byrd rule, which bars extraneous provisions from reconciliation bills and forbids any provision that increases the deficit beyond the budget window, shaped the 2017 act more than any single policy choice. On December 19, 2017, a point of order raised by Senator Sanders struck the bill’s short title, so the statute formally known as the Tax Cuts and Jobs Act is technically just Public Law 115-97; Senator Enzi’s motion to waive the rule and save the title failed 51 to 48. The parliamentarian also struck a 529-account expansion for home-schooling expenses and an endowment-tax exception for small colleges as merely incidental to the budget. Most consequentially, the rule’s ban on deficit increases beyond the ten-year window forced the drafters to sunset the individual provisions after 2025, creating the statute’s two-clock structure of a permanent corporate rate and temporary individual relief.

Q: Why did the Tax Reform Act cut corporate and individual rates at once?

Because revenue neutrality was measured on the whole package, which let each side’s priority be financed inside a single bargain. The JCT scored the individual provisions at a loss of about $122 billion and the corporate provisions at a gain of about $120 billion, for a net of roughly negative $2 billion. Cutting both rates at once was therefore not two separate tax cuts but one trade: individual rate cuts paid for by repealing individual preferences, and the corporate rate cut paid for by repealing corporate preferences such as the investment tax credit. Packaging them together gave both parties’ rate-cutting constituencies a reason to accept the preference repeals, and it gave the preference repeals a cross-party set of defenders afterward. A bill that had cut only one side’s rates would have left the other side with only the pain of base broadening and no gain.

Q: What is the sunset asymmetry between the two tax statutes?

The 1986 act had no sunsets: enacted under regular order, its provisions were permanent law until later Congresses changed them, which is why the 1990 and 1993 budget acts had to pass new legislation to raise its rates. The 2017 act, enacted through reconciliation, has two clocks. The corporate rate cut to 21 percent is permanent, as is the shift to chained CPI for indexing, while the individual rate cuts, the doubled standard deduction, the suspended personal exemption, the SALT cap, and the section 199A deduction all expire after December 31, 2025. Smaller asymmetries sit inside the business provisions: 100 percent expensing began phasing down in 2022 and phases out by 2026, and the GILTI, FDII, and BEAT provisions step up after 2025. The practical consequence is that the 2017 act’s individual relief continues only if a future Congress affirmatively extends it.

Q: What is distributional neutrality in tax reform?

Distributional neutrality is a design constraint under which a tax bill is required to leave the distribution of the tax burden across income classes roughly unchanged, as measured by JCT distributional tables. The 1986 negotiators treated it as binding: the Senate Finance background paper framed it as a requirement of the design, and provisions that violated it were renegotiated. The record shows it worked better as a negotiating rule than as a measured result, with the final distribution shifting at the margins even as the process honored the constraint. Distributional neutrality is distinct from revenue neutrality, which concerns the total, and from progressivity, which concerns the shape: a distributionally neutral bill can be paired with any level of progressivity inherited from prior law. This article takes no position on which distribution of the burden is preferable; it treats neutrality strictly as a coalition-building device.

Q: Which act changed the international tax system more?

The 2017 act, by a wide margin. The 1986 act’s international changes were modest adjustments within a system that continued to tax the worldwide income of United States companies with credits for foreign taxes. The 2017 act imposed a one time transition tax on accumulated foreign earnings under section 965, scored at a gain of about 338.8 billion dollars, moved the system toward territorial treatment of active foreign earnings, and created the GILTI, FDII, and BEAT regimes described above. Those provisions redefined how multinational income is taxed and remain among the most studied parts of the statute. The 1986 act’s enduring international legacy is smaller and indirect: its lower corporate rate reduced the incentive to shift profits abroad, but it did not rebuild the international rules the way the 2017 act did.

Q: How did the two acts treat the deduction for state and local taxes?

The 1986 act repealed the deduction for state and local sales taxes while retaining the deduction for state and local income and property taxes. The 2017 act capped the entire state and local tax deduction, income, sales, and property taxes combined, at 10,000 dollars per return, for tax years 2018 through 2025. The cap was the single largest base broadening provision in the 2017 act, scored by the Joint Committee on Taxation at a revenue gain of about 670.6 billion dollars over fiscal years 2018 to 2027. The contrast illustrates the two fiscal designs: the 1986 repeal was one of many preference eliminations that paid for rate cuts within a neutral package, while the 2017 cap was the principal offset inside a package scored as a large deficit increase. This article takes no position on which distribution of the tax burden is preferable; it records only the cap’s mechanical role as the bill’s largest individual side offset.

Q: What is the section 199A pass through deduction?

Section 199A, added by the 2017 act, allowed eligible owners of pass through businesses such as partnerships, S corporations, and sole proprietorships to deduct up to 20 percent of their qualified business income. The Joint Committee on Taxation scored it at a revenue loss of about 414.5 billion dollars over fiscal years 2018 to 2027. The provision was scheduled to expire after 2025, like the other individual side cuts. It had no equivalent in the 1986 act, which instead addressed pass through taxation indirectly through the passive loss rules and the lower individual rates that applied to pass through income. The deduction was one of the most complex provisions in the 2017 statute, with limits tied to wages paid and capital invested that phased in above income thresholds.

Q: Why was the 2017 act’s short title struck from the bill?

On December 19, 2017, Senator Bernie Sanders raised a Byrd rule point of order against the bill’s short title, the Tax Cuts and Jobs Act, on the ground that the title was extraneous matter without budgetary effect. Senator Mike Enzi offered a motion to waive the point of order, which failed 51 to 48. The title was therefore struck from the enrolled bill, leaving the statute as Public Law 115-97 without a formal short name, though the struck title remains the universal shorthand. The episode is often treated as trivia, but it demonstrates how deeply the reconciliation procedure shaped the bill: a rule designed to keep non budgetary matter out of reconciliation reached into the legislation’s name.

Q: How did the passive loss rules change tax shelters?

Before 1986, high income taxpayers routinely used passive investments, especially real estate and equipment leasing partnerships, to generate paper losses that offset wages and other ordinary income. Section 502 of the 1986 act added section 469 of the Internal Revenue Code, which generally disallowed passive activity losses except against passive income. The shelter industry that had defined high end tax planning in the early 1980s collapsed, because the mechanism it sold no longer worked. The rules continued in force long after the 1986 rates were rewritten, and IRS Publication 925 describes the regime as part of the law. The passive loss rules are the clearest example of the 1986 durability pattern: a base provision that survived because it eliminated a preference outright and its beneficiaries reorganized around the new rules.

Q: What does chained CPI indexing do to the individual tax brackets?

The 2017 act replaced the consumer price index with the chained consumer price index for the annual inflation adjustments to the tax brackets, the standard deduction, and other indexed parameters. The chained index grows more slowly because it accounts for consumers substituting toward cheaper goods when prices change. Over time, slower bracket growth pushes more income into higher brackets than the old index would have, raising revenue gradually and perpetually. Unlike the other individual provisions in the 2017 act, the chained CPI change was enacted as permanent law. It is the least visible permanent change in the statute and one of the most durable in its fiscal effect, precisely because its operation is technical and automatic instead of the subject of recurring votes.

Q: Could a future Congress combine revenue neutrality with a permanent corporate rate cut?

In principle yes, and the arithmetic shows what it would require. A permanent corporate rate cut loses revenue every year, so neutrality would demand permanent base broadening of equal size: repealing preferences, capping deductions, or adding new minimum taxes sufficient to offset the cut in every year of the window. The 2017 act’s international provisions and the state and local tax cap show the available instruments, but they covered only about half the cost of the cuts. A neutral version would have to double that offset effort or accept a smaller cut. The deeper obstacle is political instead of arithmetic. Neutrality forces the majority to eliminate preferences its own supporters value, which is the discipline no recent majority has accepted. The constraint is available to any Congress willing to pay its price.

Q: What does revenue neutrality mean in practice for a tax bill?

Revenue neutrality means the JCT scores the bill as raising roughly the same revenue over the budget window as current law would have raised, so every rate cut must be matched by an offset inside the same bill. In 1986 the mechanism was explicit: about $122 billion of individual revenue loss against about $120 billion of corporate revenue gain, a net of roughly negative $2 billion over five years. In practice the rule turns the tax-writing committees into a closed market where rate cuts are purchased with repealed preferences, which is why neutral bills broaden the base. Revenue neutrality is distinct from distributional neutrality: a bill can be revenue neutral while shifting who pays, and the 1986 negotiators treated holding the distribution fixed as a separate binding constraint. A bill that declines revenue neutrality, as the 2017 act did under its $1.5 trillion instruction, frees rate cuts from offsets but must then pass through reconciliation with its deficit discipline.