Revenue Estimation

Chapter 63 examines the empirical studies, and Chapter 64 the proposals themselves.

62.1 How the tax expenditure is measured

The statutory concept. The Congressional Budget Act defines tax expenditures as revenue losses attributable to federal tax provisions that allow a special exclusion, exemption, or deduction from gross income, or that provide a special credit, a preferential rate of tax, or a deferral of tax liability.¹ Deferral is named expressly. The difficult question is therefore not whether a deferral can be a tax expenditure but what baseline it departs from. The statute does not say, and both official estimators describe the choice as a matter of judgment.²

The Joint Committee's baseline. The staff of the Joint Committee on Taxation measures against a normal income tax. For administrative feasibility, it assumes that a normal income tax would tax capital gains in full in the year they are realized through sale, exchange, gift, or transfer at death. Deferral until realization is therefore not a tax expenditure, but a further deferral beyond the year of sale, exchange, gift, or transfer at death is.³ An exchange of property for materially different property is a realization event, and the Code's default is that realized gain is recognized.⁴ Section 1031 postpones recognition past the year of the exchange, and on the Joint Committee's baseline that postponement is the tax expenditure.

Treasury's baseline and its change of position. The Treasury's Office of Tax Analysis presents its list against two baselines, a normal tax baseline and a reference tax law baseline. Under both, income is taxable only when it is realized in exchange, and the deferral of tax on unrealized gains is not a tax expenditure.⁵ For years Treasury read that principle to keep like-kind exchanges off its list. A 2014 Office of Tax Analysis paper stated that deferral under the like-kind rules was not then considered a tax expenditure in the federal budget, and it contrasted the Joint Committee's treatment.⁶ The fiscal year 2016 list has no like-kind entry. The fiscal year 2017 list added "Deferral of capital gains from like-kind exchanges," describing current law as allowing the deferral of accrued gains on assets transferred in qualified exchanges, and the item has appeared since.⁷ The two estimators now count the same provision, although they still explain their baselines in different terms.

The method. The Joint Committee measures a tax expenditure as the difference between tax liability under present law and liability recomputed without the provision. Taxpayer behavior is held constant for this purpose. The Joint Committee's revenue estimates, by contrast, reflect anticipated behavior.⁸ Treasury publishes cash-basis estimates for every item and, for selected provisions involving deferral, present-value estimates that net the later tax payments caused by current activity. Like-kind exchanges are not among the provisions for which Treasury publishes a present value.⁹

The current numbers. The Joint Committee's December 2025 report estimates the like-kind item at $53.6 billion for fiscal years 2025 through 2029: between $2.5 billion and $2.7 billion a year for corporations and between $8.0 billion and $8.3 billion a year for individuals.¹⁰ Treasury's list published in November 2024, based on law enacted as of July 31, 2024, estimates $7.61 billion for fiscal year 2025 and $72.8 billion for fiscal years 2025 through 2034, of which $64.12 billion is attributed to individuals.¹¹

A change in composition. For fiscal year 2014 the Joint Committee had estimated the item at $5.7 billion for corporations and $3.0 billion for individuals.¹² The corporate share was the larger one then and is far smaller now. Neither estimator explains the shift, but two changes made by the 2017 Act are consistent with it. Section 1031 was limited to real property, and corporate deferral had been concentrated in vehicle exchanges: in 2012, C corporations classified in vehicles and transportation accounted for 69.7 percent of corporate deferred gain. The corporate rate also fell to 21 percent, which reduces the value of any corporate deferral.¹³

62.2 Why static estimates overstate the cost

Two different questions. A tax expenditure estimate asks how much lower this year's liability is because the provision exists, with behavior unchanged. A revenue estimate for repeal asks how much more the Treasury would collect if the provision were gone and taxpayers adjusted. The Joint Committee draws that distinction itself, and Treasury cautions that its figures do not necessarily equal the change in receipts from repeal, both because behavior would change and because provisions interact.¹⁴ The tax expenditure line is often quoted as the price of section 1031. It is not a price.

The gap in the Joint Committee's own numbers. For fiscal year 2014 the Joint Committee put the like-kind tax expenditure at $8.7 billion. In February 2014 it estimated that repealing section 1031 as part of Chairman Camp's Tax Reform Act of 2014 would raise $5.8 billion over fiscal years 2014 through 2018 and $40.9 billion over 2014 through 2023; the repeal would have applied generally to transfers after December 31, 2014.¹⁵ The figures are not strictly comparable. They were prepared on different baselines, the repeal had a delayed effective date, and it was scored inside a bill that also changed rates and cost recovery. The order of magnitude still matters: a single year's tax expenditure exceeded the first five years of estimated revenue from repeal.

Why the static figure is too high as a forecast. Three reasons recur in the literature.

First, deferral is not forgiveness. Under section 1031(d) the gain survives in the basis of the replacement property. The owner claims smaller depreciation deductions and, on a later taxable sale, recognizes the deferred gain along with any new appreciation (Chapters 35 through 37). Barker, Ling, and Petrova model the value of an exchange net of those later costs and estimate that the incremental present value of a commercial real estate exchange, compared with a taxable sale, ranges from under 1 percent to about 8 percent of the relinquished property's price.¹⁶ An earlier study by Ling and Petrova, funded by a real estate industry coalition, estimated the static present value of revenue lost from 2011 real estate exchanges at between $200 million and $3 billion, depending on assumptions.¹⁷

Second, behavior would change. Without exchanges, some owners would hold longer, and others would sell on the installment method, contribute property to an operating partnership, or not sell at all. The same authors argued in a working paper that these responses would greatly reduce the revenue a static analysis implies.¹⁸

Third, provisions interact. The value of deferral depends on rates, recovery periods, and the treatment of property at death, and Treasury notes that its estimates would differ if provisions changed simultaneously.¹⁹

Why the conclusion is narrower than it sounds. The overstatement argument is strongest as a claim about what repeal would raise. It is weaker as a claim about what the provision costs, for four reasons.

The first is death. Property acquired from a decedent takes a basis equal to its value at the date of death, so an exchange chain that ends with the owner's death converts deferral into permanent exclusion (Chapter 46).²⁰ The Joint Committee lists the exclusion of capital gains at death as a separate item, estimated at $379.3 billion for fiscal years 2025 through 2029, and the two items interact in a way neither estimator quantifies.²¹

The second is the contrary estimate. Johnson argued in 2008 that most gain not recognized in an exchange is never taxed, and he put the expected eventual tax at between 2 and 4 percent of the gain realized in an exchange. His support was research on how much accrued gain on capital assets generally is ever realized, not data on exchanges.²²

The third is the evidence most often cited for eventual taxation. The Ling and Petrova finding that 88 percent of exchange replacement properties were disposed of in taxable sales describes properties resold within the study's window. It does not measure the share held until death (63.2).²³

The fourth is discounting. Barker, Ling, and Petrova point out that the Treasury's low cost of waiting makes a deferral cheaper to the government than it is valuable to the taxpayer.²⁴ That point lowers the present-value cost to the Treasury, but only on the premise that the deferred tax is eventually paid.

The defensible statement is that static tax expenditure figures overstate the revenue repeal would raise, a proposition both estimators accept. Whether they overstate the provision's cost in present value depends on how often exchange chains end in a taxable sale rather than a basis step-up, and that question is open (63.5).

62.3 Deferral recapture in the estimating window

The mechanism. Revenue legislation is scored over a ten-year window, and the time path of a deferral provision's score inside that window has its own logic. Under present law the Treasury collects part of the deferred tax later, through smaller depreciation deductions on the replacement property and through the deferred gain recognized on an eventual taxable sale. Repeal or a cap moves the tax to the year of the exchange, but it also gives the replacement property a cost basis, which means larger depreciation deductions and a smaller gain later. Part of what repeal raises in its early years is therefore given back in later years. This treatise calls the given-back portion deferral recapture; the term is descriptive, not official. How much of it falls inside the window depends on how quickly the replacement property is recovered and resold. Five-year property returns most of the difference within the window. Residential rental property recovered over 27.5 years and nonresidential real property recovered over 39 years return it slowly, much of it after the window closes.²⁵

Three published profiles. The official scores are consistent with that mechanism, although none of them decomposes its figures.

Repeal for all property. The Joint Committee's 2014 estimate for the Camp repeal rose steadily, from $0.6 billion in fiscal year 2015 to $11.7 billion in fiscal year 2023, for $40.9 billion over the window.²⁶

Repeal for personal property only. The 2017 Act's restriction of section 1031 to real property was scored at $0.5 billion for fiscal year 2018, rising each year to $7.2 billion for fiscal year 2027, for $31.0 billion over 2018 through 2027.²⁷ Ending exchanges of short-lived equipment might be expected to produce front-loaded receipts, since the larger depreciation deductions on a cost-basis replacement catch up within a few years. The table does not say why the profile rises instead, but the rise is what one would expect from another provision of the same Act, which set additional first-year depreciation at 100 percent for qualifying property placed in service after September 27, 2017, and before 2023, stepping down by 20 percentage points a year after that.²⁸ A trade-in whose gain became taxable could often be offset by expensing the replacement, and the offset shrank as the percentage fell. Treasury economists had documented the substitution before the Act: Hertz suspended its U.S. car rental exchange program from August 2010 until October 2012 because 100 percent bonus depreciation on newly acquired vehicles more than offset the gains recognized on vehicle dispositions.²⁹

Caps in the President's budgets. Treasury's fiscal year 2016 proposal would have capped deferral on real property at $1 million per taxpayer per year and ended exchanges of art and collectibles. It was scored at $659 million in its first year and about $2.0 billion to $2.2 billion a year after that, $19.5 billion over ten years.³⁰ The fiscal year 2017 version applied the $1 million cap to all gain deferred under section 1031, not only gain on real property, and its profile is different: $2.7 billion in the first year, a peak of $7.8 billion in the second, then a decline to between about $3.7 billion and $4.0 billion a year over the last five years, $47.3 billion in total.³¹ The fiscal year 2025 proposal returned to real property, with a cap of $500,000 per taxpayer ($1 million on a joint return), and was scored at $680 million in its first year and between $1.9 billion and $2.4 billion a year after that, $19.7 billion in total.³² The Green Books do not explain the time paths. The front-loaded profile of the one proposal that reached personal property is what deferral recapture predicts, because its giveback arrives inside the window. The real-property-only proposals give back slowly, and their profiles stay level.

Reading a score. Three practical points follow.

First, a ten-year score for a real property proposal counts the early receipts in full while leaving most of the later giveback outside the window. It therefore overstates long-run revenue more than a score for short-lived property would.

Second, the scope of the property covered changes both the size and the shape of a score. The $1 million cap was the same in the fiscal year 2016 and 2017 proposals; the largest design difference between a $19.5 billion score and a $47.3 billion score was the extension beyond real property.

Third, a present-value estimate would net the giveback wherever it falls. Treasury prepares such estimates for some deferral provisions but not for this one, so a reader who wants the long-run cost of section 1031 has to build it from the assumptions discussed in 62.2.³³