Hyperlocal Responses to the SALT Deduction Limitation

Manoj Viswanathan *


The legislation commonly known as the Tax Cuts and Jobs Act (TCJA), 1 The official name is “An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018.” Pub. L. No. 115-97, 131 Stat. 2054 (2017). Though the name change to “The Tax Cuts and Jobs Act” was struck as a non-germane amendment, this truncated name persists in political and professional circles. David Kamin, David Gamage, Ari Glogower, Rebecca Kysar, Darien Shanske, Reuven Avi-Yonah, Lily Batchelder, J. Clifton Fleming, Daniel Hemel, Mitchell Kane, David Miller, Daniel Shaviro, & Manoj Viswanathan, The Games They Will Play: Tax Games, Roadblocks, and Glitches Under the 2017 Tax Legislation, 103 Minn. L. Rev. 1439, 1441 n.2 (2019) [hereinafter The Games They Will Play]. enacted in December 2017, places a $10,000 limit on the federal deduction for payments of state and local taxes (SALT). 2 I.R.C. § 164(b)(6)(B) (2017). The $10,000 limitation is in addition to the existing pre-TCJA restriction permitting the deduction of either (1) state and local income tax or (2) state and local sales tax. Id. § 164(b)(5)(A). Several states with high numbers of adversely affected taxpayers have proposed or enacted legislative workarounds to this new cap. Much has been said about these state-level responses, but there has been little analysis of local-level effects or of how local governments could similarly respond. This Essay addresses that gap by (1) statistically modeling the number of taxpayers affected by the SALT deduction limitation at a ZIP-code-by-ZIP-code (rather than state-by-state) level, and (2) proposing locality-based strategies relevant to taxpayers throughout the U.S., and not just those living in highly affected states.

A hyperlocal analysis provides several key insights. First, although blue states are generally more affected by the SALT deduction limitation, many solidly red states contain high concentrations of affected taxpayers. Focusing on statewide effects masks the extent to which large numbers of affected taxpayers exist in seemingly unaffected localities. Second, although recent SALT workarounds have focused on state income taxes, there is ample opportunity for mitigating the effects of the SALT deduction limitation by focusing instead on property taxes. This would allow cities or counties, rather than states, to implement SALT workarounds tailored to their residents. This is especially relevant for taxpayers in states without a personal income tax. Third, SALT workarounds focusing on property taxes recoup more lost SALT deductions (on a percentage basis) for middle class taxpayers than for the wealthy, possibly giving these workarounds some bipartisan appeal.

This Essay contains three parts. Part I gives an overview of the $10,000 SALT deduction limitation, the state workarounds to the limitation, and recent IRS responses. Part II describes the statistical model and presents key findings. Part III shows how local jurisdictions throughout the U.S. could create their own difficult-to-challenge SALT workarounds.

I. The SALT Deduction Limitation, State Responses, and Proposed IRS Regulations

Federal income tax deductions are allowances that enable a taxpayer to reduce the amount of their income subjected to income tax. These provisions exist to both accurately measure income (a necessary component of any tax based on income) and to promote certain policy goals. 3 See Michael H. Graetz et al., Federal Income Taxation: Principles and Policies 233 (8th ed. 2018). Many deductions are characterized as below-the-line, or itemized, deductions—taxpayers typically only take these deductions if the total of these below-the-line deductions exceed the standard deduction. 4 See generally I.R.C. § 63 (providing that taxpayers not itemizing their deductions may take the standard deduction).

The TCJA made many changes to the deductions taxpayers may take. These changes included limiting the deduction for payments of state and local taxes, which includes state income taxes, local sales taxes, and property taxes, 5 See I.R.C. § 164(a)(1)-(3), (b)(5)(A); see also supra note 2 and accompanying text. and approximately doubling the standard deduction. 6 See infra note 19 and accompanying text.

The SALT deduction limitation decreased the overall cost of the TCJA by some $500 billion and was essential for keeping the legislation within its $1.5 trillion budget. 7 See The Games They Will Play, supra note 1, at 1474-75; Bob Bryan, Republicans Just Passed a Huge Tax Reform Test by the Skin of Their Teeth, Bus. Insider (Nov. 2, 2017, 5:53 PM), (explaining the $1.5 trillion budget). Taxpayers affected by the SALT limitation are generally higher-income earners, residents of highly-taxed homes, or some combination thereof. These taxpayers are concentrated in (but not limited to) highly-taxed states and localities. 8 See infra notes 21-22 and accompanying text.

Affluent, high-tax states and localities generally vote Democratic; as such, states with high percentages of taxpayers affected by the SALT deduction are generally blue. 9 The first thirteen states/districts ranked by percentage of taxpayers affected by the SALT deduction limitation (New Jersey, Connecticut, Maryland, New York, the District of Columbia, Massachusetts, California, Virginia, Oregon, Rhode Island, Illinois, Minnesota, and Vermont) went for Hillary Clinton in the 2016 presidential election. See Presidential Election Results: Donald J. Trump Wins, N.Y. Times, (Aug. 9, 2017, 9:00 AM ET),; infra note 21 (listing the states with the highest percentage of affected taxpayers). Thus, these states have openly opposed the SALT deduction limitation with little fear of political consequences. In addition to filing suit against the Department of Treasury, 10 See Complaint for Declaratory and Injunctive Relief at 1, New York v. Mnuchin, No. 18-cv-6427 (S.D.N.Y. July 17, 2018), 2018 WL 3437090. several states have implemented legislative strategies that would mitigate the limitation’s effect on their taxpayers. 11 See Kathleen K. Wright, Reporting a State Tax Workaround on the Federal Tax Return, 2018 St. Tax Notes 297, 297-98. SALT deduction workarounds have been enacted in New York, New Jersey, Oregon, and Connecticut, and proposed (and vetoed by the Governor) in California. See Kathleen Pender, California Lags Other High-Tax States on State and Local Deduction Workaround, S.F. Chron. (May 19, 2018, 6:00 AM),; Steven M. Rosenthal, IRS Warns States To Slow Down on Charitable SALT Workarounds, TaxVox (May 24, 2018),

States with large numbers of affected taxpayers are motivated to institute workarounds for several reasons. One, state leaders implementing workarounds are able to more strongly identify as anti-Trump and his administration, which has expressive appeal; two, the workarounds have a net economic benefit for state taxpayers with little to no additional cost to the state; and three, the benefits from these workarounds could inure to both the taxpayer and the state if a less than complete credit is awarded.

The most common version of the SALT deduction workaround involves the taxpayer making a charitable contribution to the state government, with the taxpayer receiving both a state income tax credit for the donation and a federal charitable contribution deduction. 12 Unlike the SALT deduction limitation, taxpayers may deduct up to 60% of their adjusted gross income by making charitable cash contributions. I.R.C. § 170(b)(1)(G) (2017). If the state tax credit is 100% of the contribution, the taxpayer has completely circumvented the SALT deduction limitation. 13 For example, a $100 state tax credit for a federally deductible $100 charitable contribution to the state reduces a taxpayer’s state tax bill by $100 while also yielding a $100 federal deduction (assuming the taxpayer itemizes), making the $10,000 SALT deduction limitation irrelevant. Although federal charitable contributions are typically reduced by the value of any consideration received from making the contribution, judicial precedent and IRS materials (prior to August 2018) indicated that the value of state tax benefits should not trigger this “quid pro quo” reduction. 14 See Joseph Bankman et al., State Responses to Federal Tax Reform: Charitable Tax Credits, 2018 St. Tax Notes 433, 433-35 (2018).

In response to these state-sponsored charitable contribution strategies, the IRS in August 2018 proposed regulations stating that charitable contribution deductions are now “reduced by the amount of any state or local tax credit that the taxpayer receives or expects to receive in consideration for the taxpayer’s payment or transfer.” 15 Contributions in Exchange for State or Local Tax Credits, 83 Fed. Reg. 43,563, 43,571 (proposed Aug. 27, 2018) (to be codified at 26 C.F.R. pt. 1). The proposed regulations treat any significant state tax benefit obtained from making a charitable contribution as a deduction-reducing quid pro quo. Although not free from all potential legal challenges, the IRS regulations will likely curtail the various state-based SALT workaround strategies. 16 See, e.g., Comments of the Attorneys General of New Jersey, California, Connecticut, and New York, Re: IRS REG-112176-18, Proposed Rulemaking re: Contributions in Exchange for State or Local Tax Credits 83 Fed. Reg. 43563 (Aug. 27, 2018), Oct. 11, 2018 (arguing that the IRS’s proposed regulations to undermine state and local workarounds are arbitrary and capricious); Daniel Hemel, Pounding SALT?: On the Challenges Facing Treasury and the IRS in Their Effort To Rein in State Charitable Credits, Medium (Aug. 27, 2018), (arguing that the proposed regulations’ applicability to state, but not foreign, tax benefits could render them invalid). However, there is still ample opportunity for localities (cities and/or counties) to enact their own SALT workarounds outside the ambit of the IRS’s proposed regulations.

II. A Hyperlocal Model of the SALT Deduction Limitation

This hyperlocal model is the first to estimate the effects of the SALT deduction limitation for all taxpayers in tax year 2018, across all ZIP codes, and across all income groups. 17 An online version of the model is available at Zip Code Query, U.C. Hastings Ctr. on Tax L., (archived Apr. 3, 2019). Replication code and replication data are available at State Rankings of SALT Deductions, U.C. Hastings Ctr. on Tax L., (archived Apr. 9, 2019). Previous analyses of the SALT deduction limitation have focused on specific states, 18 See, e.g., State of Cal. Franchise Tax Bd., Preliminary Report on Specific Provisions of the Federal Tax Cuts and Jobs Act 6-17 (2018), (focusing solely on the effect of the SALT deduction limitation on California). or have not taken into account other important changes made by the TCJA, such as the increased standard deduction. 19 The increased standard deduction means that some taxpayers with greater than $10,000 of SALT payments will not be adversely affected by the limitation. The 2018 standard deduction is $12,000 and $24,000 for single and married filers, respectively, up from $6,350 and $12,700 in 2017. IRS, Dep’t of the Treasury, Publication 501, Dependents, Standard Deduction, and Filing Information 24 (2018),; IRS, Dep’t of the Treasury, Publication 501, Exemptions, Standard Deduction, and Filing Information 25 (2017), To create this statistical model, data from the IRS Statistics of Income were used to predict both specific numbers of taxpayers adversely affected and the extent to which these taxpayers are made worse off by the SALT deduction cap. 20 A complete description of the methodology used to create the model can be found in the Appendix.

Using the model to analyze the SALT deduction limitation’s state-level effects confirms some traditional wisdom. Ranked by percentage of affected filers, the first twelve states (and the District of Columbia) were won by Hillary Clinton in the 2016 presidential election. 21 The state with the highest percentage of affected taxpayers is New Jersey (19.8%), followed by Connecticut (19.0%), Maryland (17.2%), New York (15.7%), the District of Columbia (15.5%), Massachusetts (14.2%), California (13.9%), Virginia (11.6%), Oregon (11.3%), Rhode Island (9.8%), Illinois (9.3%), Minnesota (9.0%), and Vermont (7.4%). State Rankings of SALT Deductions, supra note 17; see Presidential Election Results, supra note 9. Next is Pennsylvania (7.0%), which went for Donald Trump. Presidential Election Results, supra note 9. Unsurprisingly, five states with existing or proposed legislative workarounds to the SALT deduction limitation are in the top nine states by percentage of affected filers, and California, New York, and New Jersey round out the top three states with the largest number of filers affected. 22 California’s 2.4 million affected taxpayers are the largest number in any one state, followed by New York (1.5 million) and New Jersey (860,000). Connecticut is tenth, with 330,000 affected taxpayers, while Oregon is seventeenth, with 209,000 affected taxpayers. See State Rankings of SALT Deductions, supra note 17.

But this state-level analysis obscures the fact that states are heterogeneous. 23 Because the SALT deduction limitation only affects wealthier taxpayers, all states, including blue ones, unsurprisingly contain areas with virtually no affected taxpayers. Many affected taxpayers live in states (and ZIP codes) that are reliably Republican. In Forsyth County, Georgia, for instance, approximately 15% of its 67,000 taxpayers will be affected by the SALT deduction limitation. 24 This number can be obtained through the data available on the online version of the model for ZIP codes 30028, 30040, and 30041, which comprise Forsyth County. See Zip Code Query, supra note 17. In the 2016 election, 71.71% of Forsyth County voted for Donald Trump. Official County Results, Ga. Election Results, (last updated Nov. 15, 2016, 12:35:21 PM EST). This is a greater percentage of affected taxpayers than in forty-six states, including California. Forsyth County is not anomalous. With the exception of Alaska, all states contain at least one ZIP code with at least 13% of taxpayers affected. 25 Alaska’s most affected ZIP code, 99516 (in Anchorage), has 8.3% affected taxpayers. See Zip Code Query, supra note 17. To the extent these affected taxpayers (in generally unaffected states) can push for legislative change, they are more likely to successfully exert political pressure on their local rather than state governments.

If the relevant local government worked with these affected taxpayers to promulgate a SALT workaround strategy focused on maintaining the full deductibility of property taxes, this strategy would recoup a significant proportion of SALT deductions that would otherwise be lost. Because a taxpayer is still permitted to deduct her first $10,000 of SALT payments, full deductibility of her property taxes will recoup a percentage of lost SALT deductions that is greater than the percentage that property taxes are of total SALT payments. For example, a taxpayer with liabilities of $10,000 of property tax and $10,000 of state income tax has a total SALT liability of $20,000. This taxpayer will recoup 100% of her lost SALT deductions with full property tax deductibility, even though property taxes are only 50% of her total SALT liability. Across all taxpayers, income groups, and ZIP codes, full deductibility of property taxes results in recouping approximately 60% of SALT deductions lost, even though property taxes are only 34% of SALT payments deducted pre-TCJA. 26 This estimate comes from my model. In contrast, my model demonstrates that obtaining full deductibility for either income taxes or sales taxes (whichever is greater) would recoup approximately 88% of SALT deductions lost.

Because property taxes are a smaller proportion of total SALT liability for wealthier homeowners, the percentage of SALT deductions recouped by full deductibility of property taxes varies by income class. For taxpayers with adjusted gross incomes (AGIs) of less than $200,000, full deductibility of property taxes results in restoring approximately 75% of deductions lost by the SALT deduction limitation. 27 This estimate comes from my model. Approximately 59% of all taxpayers affected by the SALT deduction limitation have AGIs of less than $200,000, according to my model. For taxpayers with AGIs of less than $100,000, full deductibility of property taxes restores approximately 94% of SALT deductions lost. 28 This estimate comes from my model. Approximately 10% of all taxpayers affected by the SALT deduction limitation have AGIs of less than $100,000, according to my model. For taxpayers with AGIs greater than $200,000, full deductibility of property taxes results in recouping only about 38% of SALT deductions lost. 29 This estimate comes from my model.

Full deductibility of property taxes is a SALT workaround that benefits middle class taxpayers more (by percentage of lost SALT deductions recouped) than wealthier taxpayers. As a result, such a workaround is more defensible on distributional grounds than a full repeal of the SALT deduction limitation. This, combined with the fact that taxpayers in red states could also easily participate, could result in bipartisan tolerance—if not outright support—for these SALT workarounds based on property tax.

Maintaining deductibility of property taxes is even more valuable for the 910,000 affected taxpayers living in the seven states with no personal income tax. 30 The seven states with no personal income tax are Alaska, Florida, Nevada, South Dakota, Texas, Washington, and Wyoming. See Morgan Scarboro, Tax Found., State Individual Income Tax Rates and Brackets for 2018, at 4-9 (2018). The 910,000 figure comes from my model. See State Rankings of SALT Deductions, supra note 21. Instead of deducting state and local income taxes paid, these taxpayers deduct sales taxes paid. 31 See I.R.C. § 164(b)(5)(A) (2017). Taxpayers may deduct either sales taxes paid or state income taxes paid, but not both. See supra note 2. These filers usually pay significantly more in property taxes than they do in sales taxes. As a result, full deductibility of property taxes for these filers would result in recouping approximately 95% of SALT deductions. 32 This estimate comes from my model. In contrast, full deductibility of sales taxes paid would recoup only about 49% of SALT deductions lost. 33 This estimate comes from my model.

III. Local SALT Workarounds

As discussed in Part I, the state-sponsored charitable contribution workarounds are stymied by the proposed IRS regulations because the state tax benefits awarded reduce the tax liability otherwise owed by a specific taxpayer. Thus, the tax benefits received have a readily ascertainable value. The quid pro quo of the charitable contribution is readily identifiable, and the taxpayer’s charitable contribution is then reduced correspondingly.

Tax benefits conferred more broadly, however, cannot easily be treated as quid pro quo even if these tax benefits are funded by charitable contributions. For example, if California receives an unexpected number of unrestricted charitable contributions and chooses to lower state income tax rates because of these unexpected funds, no conceivable quid pro quo has occurred. Making contributions to a nongovernmental tax-exempt organization (rather than the state) makes quid pro quo characterization even more unlikely—if a state earmarks tax revenues for specific beneficiaries to which the tax-exempt organization could instead make grants, the state could broadly eliminate taxes (as opposed to crediting specific individuals) if the funding for those beneficiaries is guaranteed.

Although property tax regimes vary widely across jurisdictions, amounts owed are typically known prior to the close of the taxable year. 34 See, e.g., Property Tax Due Dates, N.Y.C. Dep’t of Fin., (archived Mar. 25, 2019) (“Bills are generally mailed and posted on our website about a month before your taxes are due.”). In California, for example, property taxes cannot increase (absent a sale) by more than a fixed percentage every year. 35 Proposition 13 amended California’s constitution to achieve this result. See Cal. Const. art. XIII A, § 2(b). A homeowner in San Francisco knows what her 2018 property tax bill will be well before the year ends and, reciprocally, San Francisco knows the taxes it can expect to receive. 36 Although assessment values in California may change at sale, this transfer rate is incorporated into San Francisco’s budget. See Office of the Controller, City & Cty. of S.F., FY 2018-19 and FY 2019-20 Revenue Letter: Controller’s Discussion of the Mayor’s Proposed Budget 13 (2018) (factoring transfer rate into property tax revenue estimates).

Consider a group of homeowners in San Francisco all expecting to pay more than $10,000 in California state income taxes. 37 Assume that all these taxpayers will also be itemizing and not taking the standard deduction. Because these homeowners have already reached the $10,000 SALT deduction limit, no portion of their property tax payments will be deductible for federal income tax purposes. These homeowners can approach San Francisco and request that the property tax levied on their properties be reduced to zero provided that the homeowners raise, via charitable contributions, a threshold amount equal to, say, 110% of the property tax revenue that would have otherwise been collected. 38 Setting the threshold amount to greater than the property taxes previously collected ensures that San Francisco collects even more money than it had previously. Due to the full deductibility of charitable contributions, this is still better for taxpayers than paying a smaller amount of nondeductible property taxes. If San Francisco agrees, the homeowners could form an LLC whose organizational documents specify that a charitable contribution to San Francisco will be made only if the LLC reaches the threshold level of funding. If this target is not met, the monies would be returned to the homeowners. If San Francisco agrees, and the requisite funding is raised, San Francisco would have eliminated property taxes not for any individual taxpayer, but for a group of property owners.

It would not be necessary for localities wishing to implement SALT property tax workarounds to obtain buy-in from all property tax payors. The locality could divide the property tax base into smaller groupings chosen by homeowners, and eliminate (or simply reduce) property taxes for the groupings that have raised the requisite amount of charitable contributions. The groupings would most readily consist of property owners with similar incomes and property tax obligations in geographic proximity to one another. Additionally, as illustrated below, economic viability of the SALT property tax workaround is not even dependent on unanimous participation within a group—less than 100% participation could still result in lowered tax obligations for the group as a whole even if, contrary to expectations, the lowered tax obligations are treated as a quid pro quo.

Consider a group of five identical taxpayers, who each expect to owe $15,000 of property tax that is not deductible because the $10,000 SALT cap has already been reached. Assume further that their marginal tax rate is 40%. Each taxpayer should be willing to charitably contribute up to $25,000 if the government agreed to eliminate their property tax liability. If each taxpayer charitably contributed $25,000, their taxable income would be reduced by that amount, resulting in tax savings of $10,000. 39 That $25,000 in taxable income would have been taxed at a 40% rate, resulting in a tax payment of $25,000 x 40%, or $10,000. Because each taxpayer would be making a charitable contribution of $25,000, but only receiving tax savings of $10,000, the cost to each taxpayer would be $15,000. In other words, by making a charitable contribution of $25,000 and having their property taxes eliminated, the taxpayer would be out $15,000—the same amount they would otherwise owe in property taxes. To the taxpayer, it should make little difference that they pay $15,000 by making a charitable contribution and having their property taxes eliminated, versus paying $15,000 in property taxes.

What if only three of the five taxpayers charitably contributed $25,000 each, and in exchange for this contribution, the government relieved all five taxpayers of their $15,000 property tax obligation? (The government should in theory be willing to do so, as it would receive the same total amount of funds at the end of the day: $75,000 in charitable contributions from three taxpayers paying $25,000 each, versus $75,000 in property taxes from five taxpayers paying $15,000 each.) The resulting property tax elimination could at most be seen as a $15,000 quid pro quo for the three contributors. Even if the $15,000 property tax elimination was treated as a quid pro quo, the deductible amount of the charitable contribution amount would be reduced by that amount, from $25,000 to $10,000, for only the contributing taxpayers. While the full $25,000 charitable contribution amount would have resulted in tax savings of $10,000, the reduced $10,000 charitable contribution amount would result in tax savings of only $4,000. 40 The $25,000 charitable contribution would reduce taxable income by $25,000, resulting in tax savings of $25,000 x 40%, or $10,000. By contrast, a $10,000 charitable contribution would reduce taxable income by $10,000, resulting in tax savings of $10,000 x 40%, or $4,000. The additional $6,000 incurred by each of these three taxpayers for eliminating their property tax liability ($18,000 total) could easily be compensated by the two remaining taxpayers paying $9,000 each ($18,000 total). 41 From the perspective of the two original non-contributors, this would result in the same outcome as if their property taxes were paid and fully deductible. If each of the two taxpayers’ $15,000 in property taxes were eliminated, but they paid $9,000 each to the other three taxpayers, their total costs for satisfying their tax liability would be $9,000 each. Similarly, if their property taxes were paid and fully deductible, the taxpayers would have paid $15,000 in property taxes, minus $6,000 in tax savings, for a net after-tax cost of $9,000. (With a 40% marginal tax rate, a deduction of $15,000 is worth $15,000 x 40%, resulting in tax savings of $6,000.) Although safeguards to prevent free-riders could easily be implemented (by, for example, requiring each homeowner to contribute a specified amount before the charitable contribution is made) this example shows that such free-riding would not destroy the efficacy of the workaround even if, contrary to expectations, the IRS treated the property tax elimination as a quid pro quo.

The organizational structure of the property tax workaround could take many forms other than homeowners forming an LLC. Taxpayers could form a tax-exempt organization whose mission is to make periodic disbursements to San Francisco, or form a tax-exempt supporting organization whose charitable mission is to fund the same organizations supported by the property tax. Any of these structures, coupled with a property tax regime where property tax is only collected to the extent that the charitable contributions do not fully equal the property taxes previously levied by the locality, could result in tax treatment for taxpayers equivalent to deducting most, if not all, property tax payments. 42 Although tax planning costs for the first group of taxpayers seeking a workaround might be burdensome, these costs need only be borne once. Subsequent taxpayers could easily piggyback off the tax planning of earlier groups.

Because the August 2018 proposed regulations treat state tax benefits received by individuals as quid pro quos that reduce deductible charitable contributions, 43 Contributions in Exchange for State or Local Tax Credits, 83 Fed. Reg. 43,563, 43,571 (proposed Aug. 27, 2018) (to be codified at 26 C.F.R. pt. 1). characterization as a quid pro quo is more likely if a single taxpayer directly receives tax relief in exchange for her charitable contribution. However, it would be difficult for the IRS to characterize as a quid pro quo the elimination of property taxes for a group of taxpayers who have, in the aggregate, raised a specified amount of dollars via charitable contributions. Even with a broad reading of what qualifies as a quid pro quo, the locality’s decision not to tax does not create an easily cognizable benefit inuring to any individual taxpayer.


Although blue states are generally more affected by the TCJA’s $10,000 SALT deduction limitation, concentrated pockets of affected taxpayers exist throughout the U.S. Through locally sponsored SALT workaround strategies that seek to maintain deductibility of property taxes, these taxpayers could effectively mitigate the adverse effect of the TCJA’s SALT deduction cap.

There are many ways beyond those described in this essay that a SALT workaround based on property taxes could be structured, and just as many other ways the IRS might attempt to challenge them. This Essay’s goal is not to provide a blueprint by which an unassailable SALT workaround could be created, or even take a position on the normative validity of such strategies. Rather, its goal is to show that such workarounds could succeed in places and for taxpayers not previously considered, and that such workarounds could be both relatively straightforward to implement and difficult to prevent.

Appendix: Modeling the Hyperlocal Effects of the SALT Deduction Limitation

A. Basic Assumptions

This statistical model determines the extent to which taxpayers are worse off in 2018 due to the limitation of the SALT deduction cap to $10,000. All other relevant provisions of the TCJA other than the SALT deduction limitation are assumed extant, including the increased standard deduction across filing statuses and the reduction in types of expenditures qualifying as itemized deductions. 44 In 2018, the standard deduction for taxpayers filing as (1) single, (2) head of household, and (3) married filing jointly is $12,000, $18,000, and $24,000, respectively—approximately double pre-TCJA amounts. See supra note 19. The model assumes no change in taxpayer behavior between 2015 (the year from which IRS data is used) and 2018.

The model also assumes that any filer itemizing under 2018 law would also itemize under pre-2018 law; in other words, that 2018 itemizers are a subset of 2015 itemizers. This assumption is based on the increase in standard deduction amounts, reduction in permitted itemized deductions, and relative paucity of itemized deductions with greater eligibility in 2018 compared to 2015. 45 But see I.R.C. § 213(f)(2) (2017) (lowering the threshold for the medical expenses deduction from 10% to 7.5%). The expected result is that fewer taxpayers will itemize in 2018 compared to previous years.

The model does not take inflation from 2015 to 2018 into account. The income and deduction items reported in 2015 are assumed to be of equal value in 2018. Additionally, other than changes due to the increased standard deduction amounts, the model assumes that taxpayers in 2018 do not change their behavior in response to differences between 2018 and 2015 law.

B. The IRS Data File

IRS Form 1040 line item data are available from the IRS Statistics of Income 2015 ZIP Code data. 46 See SOI Tax Stats—Individual Income Tax Statistics—2015 ZIP Code Data (SOI), IRS, (last updated Aug. 21, 2018). This data file groups taxpayers by both ZIP code and adjusted gross income (within one of six ranges) 47 The adjusted gross income ranges are (1) $1 to $25,000, (2) $25,000 to $50,000, (3) $50,000 to $75,000, (4) $75,000 to $100,000, (5) $100,000 to $200,000, and (6) $200,000 or more. and provides both the total number of returns and aggregate amounts for most line items contained in IRS Form 1040 and its associated variants. 48 For our purposes, relevant categories include total number of returns, number of returns by filing status, number and aggregate amount of itemized deductions in total, number and aggregate amount of state and local income taxes, number and aggregate amount of state and local general sales taxes, number and aggregate amount of real estate taxes, number and aggregate amount of all taxes paid, number and aggregate amount of mortgage interest paid, and number and aggregate amount of charitable contributions. Although the data provide the total number of applicable returns and the aggregate amounts for each reported line item, information about individual returns is not determinable. For instance, in ZIP code 35004 (the majority of which is in Moody, Alabama) in the $25,000 to $50,000 AGI range, there are 1,350 returns, 240 of which report having taxable interest. The aggregate amount of taxable interest reported by these 240 taxpayers is $140,000. 49 SOI Tax Stats, supra note 46 (to locate, click the “Alabama” link). Although the average amount of taxable interest per these 240 returns can be calculated ($583), no additional information about the distribution of taxable interest amongst those 240 returns is available.

Additionally, unless known by operation of law, the data file does not allow for determining which filers are common to each category. For example, 100 of the 1,350 filers described above also reported receiving qualified dividends. But we cannot directly determine how many of these 100 returns also reported taxable interest. Some categories, however, are known by operation of law to be subsets of other categories. Because mortgage interest can only be deducted by taxpayers not taking the standard deduction, it can be assumed that the 280 filers in ZIP Code 35004 with AGIs between $25,000 and $50,000 paying mortgage interest are also amongst the 370 filers listed as itemizers. 50 See I.R.C. §§ 63(a), 67(b)(2), 163(h)(3) (permitting deduction of home mortgage interest only if taxpayers elect to itemize).

The IRS takes certain measures to avoid disclosure of taxpayer-specific information. All “number of returns” entries are rounded to the nearest ten. 51 IRS, Zip Code Data: Tax Year 2015 Documentation Guide 2 (n.d.). ZIP codes with less than 100 returns and those identified as a single building or nonresidential ZIP code are categorized by the IRS as “other,” and are lumped into a “99999” tag for each state. 52 Id. Categories with less than twenty returns for a particular AGI class are combined with another AGI class within the same ZIP code. 53 Id.

C. Methodology

Although the data file reports filing status data (e.g., married filing jointly, single, and head of household) for each income band within each ZIP code, the data contained in other categories is not subdivided by filing status. As such, for each income band a hybrid standard deduction was calculated and used to represent the average taxpayer within the income band. Within each income band, the probability of this average taxpayer itemizing in 2018 was determined by calculating the subset of 2015 itemizers that would continue to itemize even with the increased 2018 standard deduction amounts. This subset was determined by assuming, for each fixed-size income band, 54 The first five income bands are fixed-size income bands; the last income band, which comprises adjusted gross incomes $200,000 or greater, is of indefinite size. that the itemized deductions taken within that band would be normally distributed around the mean amount of itemized deductions for that band. By assuming a reasonable standard deviation, 55 Given a range and median, standard deviation can be estimated using range divided by six. Stela Pudar Hozo et al., Estimating the Mean and Variance from the Median, Range, and the Size of a Sample, 5 BMC Med. Res. Methodology 1, 4 (2005). the percentage of itemizers exceeding the 2018 hybrid standard deduction amount can be estimated.

Because the largest income band consists of taxpayers with adjusted gross incomes of $200,000 or greater (with no cap on the maximum amount), assuming a normal distribution for the relevant categories of income and deduction is inaccurate. To properly capture the high-end tail common to income distributions, a log-normal distribution was used. Such a distribution both limited the erroneous predictions of itemized deduction amounts lower than the hybrid 2018 standard deduction amount and more accurately modeled the tail-end itemized deduction amounts found in this uncapped income band.

Similar to how the percentage of 2015 itemizers who would still itemize under 2018 law can be determined, the subset of 2015 taxpayers paying greater than $10,000 in state and local taxes can also be estimated for each income band within each ZIP code. Because the $10,000 maximum SALT deduction is constant across all filing types, it is not necessary to create a hybrid maximum SALT deduction. Assuming that the probability of 2015 itemizers itemizing under 2018 law and the probability of 2015 itemizers paying more than $10,000 in SALT are independent events, 56 It is likely that these two events are not completely independent since a larger amount of itemized deductions probably implies greater SALT payments. But an increase in itemized deductions could also be due to other itemized deductions, such as the mortgage interest deduction. the overall probability of a taxpayer both itemizing and paying more than $10,000 in SALT can be estimated by simply multiplying these two probabilities.

Determining the SALT deduction recouped in 2018 if each taxpayer was able to deduct her property taxes in full requires knowing taxpayer data to a specificity not provided by the IRS Statistics of Income. The SALT deduction permits taxpayers to deduct either state income taxes or sales taxes. As such, the aggregate data given do not readily permit calculating which option each specific taxpayer within each ZIP code/income band will select. However, a reasonable estimate can be determined by assuming that all taxpayers within each ZIP code/income band deduct the larger category. From this an estimate for income (or sales) taxes and property taxes paid can be estimated, along with the tax savings from a full property tax deduction. With this information an estimate for the percent of SALT deductions previously lost but recouped under a full property tax deduction can be calculated for all ZIP codes/income bands. The same approach can be used to determine the percentage of SALT deductions recouped assuming full deductibility of income (or where applicable, sales) taxes.

* Associate Professor of Law, University of California, Hastings College of the Law.