April 25, 2017
Note from the Editor: This article discusses the parsonage allowance, whereby the value of a minister’s home is exempted from federal income tax. It argues that the allowance is constitutional under the Establishment Clause, and indeed desirable pursuant to important Establishment Clause values. The Federalist Society takes no positions on particular legal and public policy matters. Any expressions of opinion are those of the authors. Whenever we publish an article that advocates for a particular position, as here, we offer links to other perspectives on the issue, including ones opposed to the position taken in the article; those can be found under "related content." As always, we also invite responses from our readers. To join the debate, please email us at email@example.com.
For almost a century, Congress has excluded the value of a minister’s home from federal income tax. The Internal Revenue Code provides: “In the case of a minister of the gospel, gross income does not include (1) the rental value of a home furnished to him as part of his compensation; or (2) the rental allowance paid to him as part of his compensation,” within certain limits. Because many ministers have traditionally lived in church-owned housing, or “parsonages,” this statute—§ 107 of the Internal Revenue Code—is often called the “parsonage allowance.”
Over the last several years, some academics and litigants have attacked this longstanding tax provision as a violation of the Establishment Clause. But their arguments often fail to consider the parsonage allowance’s historical and statutory contexts, both of which show that the government has regularly adapted tax principles to the unique circumstances of religious organizations in order to promote the important Establishment Clause values of church autonomy, non-entanglement, and non-discrimination. This article will explore those contexts in order to demonstrate why the parsonage allowance is not only permissible under the Establishment Clause, but desirable.
I. The Parsonage Allowance and the Establishment Clause
The Establishment Clause provides that “Congress shall make no law respecting an establishment of religion.” To interpret this clause, the Supreme Court has employed various tests.
In some cases, the Court has applied the Lemon test, which asks whether the government’s action 1) has a religious “purpose,” 2) has the “primary effect” of “advancing” or “endorsing” religion, and 3) fosters “excessive [government] entanglement with religion.” This test has been heavily criticized by courts and commentators alike and has not been applied by the Supreme Court in a merits decision in over 12 years. At least eight current or recent Justices have called for its rejection. And in recent cases, the Court has treated the Lemon factors as “no more than helpful signposts,” if it has applied them at all.
Instead, the Court has increasingly focused on the historical meaning of the Establishment Clause and the practices that have long been permitted under it. It has also decided two prominent cases in the tax context—Walz v. Tax Commission and Texas Monthly, Inc. v. Bullock. In these cases, although the Court mentioned some of the Lemon factors, its analysis was not driven by a three-factor test. Rather, the Court focused on the history of the Establishment Clause, the nuances of the tax code, and principles unique to the tax context.
In the case of the parsonage allowance, while lower courts may feel compelled to consider the Lemon factors, it is also crucial that they consider the historical meaning of the Establishment Clause, the practices that were permitted under it, and the Court’s analyses in Walz and Texas Monthly. As explained below, the parsonage allowance is not only permissible under the Establishment Clause, but desirable, because it furthers the core Establishment Clause values of neutrality, non-discrimination, and non-entanglement. It is fully consistent with the historical meaning of the Establishment Clause. It is fully consistent with the controlling concurrence in Texas Monthly. And it is fully consistent with the plurality’s more stringent test in the same case. Finally, striking down the parsonage allowance would threaten scores of other provisions in the federal and state tax codes.
II. The Parsonage Allowance Is Consistent With a Historical Understanding of the Establishment Clause
In its most recent Establishment Clause decision, the Supreme Court reaffirmed that “the Establishment Clause must be interpreted by reference to historical practices and understandings.” It engaged in a thorough review of legislative prayer practices “[f]rom the earliest days of the Nation” that have “long endured,” and “become part of our heritage and tradition,” concluding that the “prayer practice in the town of Greece fits within the tradition long followed in Congress and the state legislatures.”
But this is nothing new; history has always been highly relevant in the Supreme Court’s Establishment Clause jurisprudence. In Marsh v. Chambers, the Court upheld a state’s practice of paying a chaplain who led legislative prayer because similar practices were “deeply embedded in the history and tradition of this country.” The history “sheds light not only on what the draftsmen intended the Establishment Clause to mean, but also on how they thought that Clause applied to the practice authorized by the First Congress.” Similarly, in Hosanna-Tabor Evangelical Lutheran Church & School v. E.E.O.C.—the Court’s first decision involving the ministerial exception, which is rooted in the Establishment Clause—the Court examined the history of colonial “[c]ontroversies over the selection of ministers,” as well as “two events involving James Madison,” to determine that “[t]he Establishment Clause prevents the Government from appointing ministers.” And in Van Orden v. Perry, the Court upheld Texas’ Ten Commandments display, with a plurality applying an analysis “driven both by the nature of the monument and by our Nation’s history.”
It should come as no surprise, then, that the Court has similarly applied a historical analytical framework in tax cases. In Walz, the Court rejected an Establishment Clause challenge to New York’s property tax exemption for church property. The Court held that “[t]here is no genuine nexus between tax exemption and establishment of religion.” It reached this conclusion based on more than two centuries of “our history and uninterrupted practice” showing that “federal or state grants of tax exemption to churches were not a violation of the Religion Clauses of the First Amendment.” In his concurrence, Justice Brennan similarly looked to the “history” and “practices of the Nation,” finding that “[t]he existence from the beginning of the Nation’s life of a practice . . . is a fact of considerable import” in determining constitutionality under the Establishment Clause. Given this “uninterrupted” and “historic practice,” Justice Brennan observed that religious tax “exemptions were not among the evils that the Framers and Ratifiers of the Establishment Clause sought to avoid.”
So what does history have to say about the tax treatment of churches and ministers? At the time of the Founding, an establishment of religion consisted of one or more of several key elements, all involving state coercion to participate in religious activity: 1) government control of the doctrine and personnel of the church, 2) government coercion of religious beliefs and practices, 3) government assignment of important civil functions to the church, and 4) government financial support of the church. The “financial support” that amounted to an establishment took very specific forms: government land grants to the established church, direct grants from the public treasury, and compulsory taxes or “tithes” for the support of churches and ministers.
By contrast, tax exemptions like the parsonage allowance were never considered to be establishments in the Founding era. As the Court said in Walz, a tax exemption “is not sponsorship since the government does not transfer part of its revenue to churches but simply abstains from demanding that the church support the state.” Far from creating an impermissible unity of church and state, a tax exemption “restricts the fiscal relationship between church and state, and tends to complement and reinforce the desired separation insulating each from the other.”
Over 200 years of unbroken history confirm that religious tax exemptions are fully consistent with the historical meaning of the Establishment Clause. Religious tax exemptions permeate state and federal tax codes, and have done so since the Founding. For example, in 1799, Virginia took steps to disestablish the Anglican Church, repealing measures that had given property to the church, and condemning them as being “inconsistent with the principles of the constitution, and of religious freedom, and manifestly tend[ing] to the re-establishment of a national church.” Yet even after it formally disestablished the Anglican Church, Virginia consistently exempted the property of “any college, houses for divine worship, or seminary of learning” from taxation. “It may reasonably be inferred that the Virginians did not view the exemption for ‘houses of divine worship’ as an establishment of religion.” The municipal government of the District of Columbia exempted “houses for public worship” from property taxes in 1802. Significantly, “[a]ll of the 50 States provide for tax exemption of places of worship, most of them doing so by constitutional guarantees.” And “[f]or so long as federal income taxes have had any potential impact on churches—over 75 years [to the 1970 case]—religious organizations have been expressly exempt from the tax.”
While property tax exemptions for churches have often included other non-profit charitable organizations as well, many other religious tax exemptions have not. Early Congresses viewed religious tax exemptions as consistent with the Establishment Clause even when the exemptions did not also apply to secular groups. For example, Congress refunded import duties paid by religious organizations on religious articles like plates for printing Bibles, church vestments, furniture, and paintings, and church bells; it also exempted all churches and appurtenant property in D.C. “from any and all taxes or assessments, national, municipal, or county.” Similarly, “[a]t least 45 States provide exemptions for religious groups without analogous exemptions for other types of nonprofit institutions.” These exemptions range from sales and beverage tax exemptions for sacramental wine and meals served by churches to sales tax exemptions for church vehicles used to transport people for religious purposes. And, analogously to the federal tax code’s § 107, numerous states exempt clergy housing from taxation and have done so for many decades.
The distinction between these permissible religious tax exemptions and prohibited government sponsorship of religion is not mere formalism or historical accident. Exempting religious actors from taxation is qualitatively different from providing direct financial support because tax exemptions respect First Amendment values by protecting church autonomy and reducing government entanglement with religion. The Supreme Court “has long recognized that the government may (and sometimes must) accommodate religious practices and that it may do so without violating the Establishment Clause.” Thus, it is a “permissible legislative purpose to alleviate significant governmental interference with the ability of religious organizations to define and carry out their religious missions.”
Imposing additional taxes on ministers’ housing allowances would interfere with the ability of churches to carry out their religious missions by diverting scarce resources away from their core First Amendment activities. As Justice Brennan recognized in Walz: “[T]axation would surely influence the allocation of church resources,” with “public service activities . . . bear[ing] the brunt of the reallocation.” And taxation “would bear unequally on different churches, having its most disruptive effect on those with the least ability to meet the annual levies assessed against them.”
The parsonage allowance not only alleviates a government-imposed burden on churches, but also reduces government entanglement in religion by avoiding the “direct confrontations and conflicts” between ministers and the government that would occur without it. With increased taxation come more IRS deficiency actions, more “tax liens, [and] tax foreclosures.” Religious tax exemptions thus “constitute a reasonable and balanced attempt to guard against” the “latent dangers inherent in the imposition of . . . taxes.”
In short, while the Establishment Clause prohibits the types of direct financial support that prevailed in colonial establishments—land grants, direct grants from the treasury, and compulsory “tithes” to support churches and ministers—it does not bar the tax exemption for parsonages. Such exemptions were common at the time of the Founding and actually further the core Establishment Clause goals of alleviating government burdens on religion, avoiding discrimination among churches, and avoiding entanglement between church and state.
III. The Parsonage Allowance Is Consistent With the Controlling Opinion in Texas Monthly
The parsonage allowance is also consistent with the Supreme Court’s decision in Texas Monthly. Nearly 20 years after a 7–1 majority in Walz upheld tax exemptions for churches as a practice “deeply embedded in the fabric of our national life,” a fractured Court in Texas Monthly invalidated a sales tax exemption that applied exclusively to “periodicals . . . that consist wholly of writings promulgating the teaching of [a] faith” and “books that consist wholly of writings sacred to a religious faith.” No opinion received more than three votes.
Justice Brennan, in a plurality opinion joined by Justices Marshall and Stevens, concluded that the sales tax exemption violated the Establishment Clause because it constituted a “subsidy exclusively to religious organizations,” “burden[ed] nonbeneficiaries markedly,” and “c[ould] not reasonably be seen as removing a significant state-imposed deterrent to the free exercise of religion.” He argued that a religious tax exemption would be constitutional only if it were part of a broader scheme that provided benefits to “a large number of nonreligious groups as well.”
Justice White concurred in the result, but avoided the Establishment Clause altogether. He concluded that the sales tax exemption “discriminates on the basis of the content of publications,” and therefore “is plainly forbidden by the Press Clause of the First Amendment.”
Justice Blackmun, joined by Justice O’Connor, concluded that the sales tax exemption violated the Establishment Clause, but offered a “narrow resolution of the case.” Specifically, they acknowledged that the exemption might be upheld if it were coupled with an exemption for “philosophical literature” covering similar topics, and that “the Free Exercise Clause [may even] require a tax exemption for the sale of religious literature by a religious organization.” But they reasoned that “by confining the tax exemption exclusively to the sale of religious publications, Texas engaged in preferential support for the communication of religious messages.” This sort of “statutory preference for the dissemination of religious ideas offends our most basic understanding of what the Establishment Clause is all about.” Thus, for the Blackmun/O’Connor concurrence, the critical issue was that the tax exemption applied “exclusively” to religious literature, and that this had the effect of giving preferential support to religious messages. This focus on a preference for religious messages was the “narrowest grounds” for decision, and is therefore the controlling opinion under Marks v. United States.
The parsonage allowance is distinguishable from the tax exemption struck down in Texas Monthly in important ways. First, unlike Texas Monthly, where the tax exemption for religious literature stood alone, the parsonage allowance is one of many tax exemptions for housing allowances, most of which are nonreligious. These include exemptions for any employee who receives lodging for the convenience of his employer, any employee living in a foreign camp, any employee of an educational institution, any member of the uniformed services, any government employee living overseas, any citizen living abroad, and any employee temporarily away from home on business. It is as if, in Texas Monthly, the state had coupled the tax exemption for religious literature with a tax exemption for business literature, scientific literature, educational literature, travel literature, and government literature. That would not be a form of “preferential support” for religious messages; it would be a form of putting religious messages on the same footing as many other secular messages deemed socially beneficial. Indeed, in such circumstances, Justices Blackmun and O’Connor would likely have argued that “the Free Exercise Clause requires a tax exemption for the sale of religious literature.”
Second, the Blackmun/O’Connor concurrence did not address preferential support for “religion” generally; instead, it emphasized that the Court was dealing with “the taxation of books and journals,” which implicates “three different Clauses of the First Amendment: the Free Exercise Clause, the Establishment Clause, and the Press Clause.” Accordingly, its Establishment Clause analysis placed great weight on the fact that the tax exemption applied specifically to religious “literature”—mentioning this point, or some variation of it, no less than eighteen times. Of course, the parsonage allowance applies to housing, not religious literature. And it applies regardless of whether the minister who lives there is involved in spreading a religious message. In that sense, because it is tied to property, the parsonage allowance is much more like the property tax exemption upheld in Walz. Indeed, while some ministers certainly use their homes to teach and counsel their congregations, the connection between ministers’ homes and religious messages is even weaker than the connection between actual church buildings and religious messages in Walz. And the Blackmun/O’Connor concurrence certainly did not disturb Walz’s ruling on exemptions for churches more generally.
IV. The Parsonage Allowance Satisfies Even the More Restrictive Test of the Texas Monthly Plurality
Even assuming the Texas Monthly plurality is controlling, the parsonage allowance still satisfies its more stringent test. Under the Texas Monthly plurality, “[w]hat is crucial is that any subsidy afforded religious organizations be warranted by some overarching secular purpose that justifies like benefits for nonreligious groups.” The fit between the overarching secular purpose and the benefit for religious organizations need not be perfect. Rather, it is enough if “it can be fairly concluded that religious institutions could be thought to fall within the natural perimeter [of the legislation].” Because this test is even more stringent than the Lemon test, a statute that satisfies the Texas Monthly concurrence satisfies Lemon as well.
Section 107(2) is part of a broad scheme of tax exemptions serving the same secular purpose: ensuring fair tax treatment of employee housing costs. Since its inception, the federal income tax system has recognized that some housing costs are incurred primarily for “the convenience of the employer”—not for the employee’s personal consumption—and are therefore not income. Many tax provisions embody this doctrine. Some provisions demand case-by-case analysis of each situation, but others establish bright lines for certain classes of workers, reducing the disputes and non-uniformity that would result from an individualized, case-by-case approach. This reduction of disputes and non-uniformity is especially vital in the context of ministers, because it fulfills the Establishment Clause’s core directives of limiting entanglement between church and state and avoiding discrimination among religious groups.
A. Non-Ministers Receive a Variety of Tax-Exempt Housing Benefits Under the “Convenience Of The Employer” Doctrine
The parsonage allowance codified in § 107(2) is part of a broad package of tax exemptions that all trace their origin to the “convenience of the employer” doctrine, which is as old as the federal income tax itself. One cannot understand § 107(2) without understanding the convenience of the employer doctrine—including its rationale, its history, and its codification throughout the tax code.
1. Rationale of the Doctrine
The convenience of the employer doctrine flows from a very basic principle about the nature of income: for something to qualify as income, there “must be an economic gain, and this gain must primarily benefit the taxpayer personally.” For example, a worker might receive any number of things that simultaneously benefit her and her employer’s business—such as meals, travel, entertainment, and office furnishings. But if these things are primarily intended to further the business of the employer, rather than compensate the employee, they are not treated as income.
The same principle applies to lodging. In general, when an employee receives ordinary lodging or a housing allowance, it does not benefit the employer other than by compensating the employee, and so the value of the lodging is treated as income. But in some cases, the lodging is provided primarily “for the convenience of the employer.” Common examples include hotel managers who must live at the hotel, military officers who must live in the barracks, or commercial fishermen who must live on a ship. For these workers, the lodging is an important component of their job. As one early court put it, it is “part of the maintenance of the [employer’s] general enterprise,” not “part of the individual income of the laborer.” In such cases, excluding the lodging from income does not confer a special benefit; rather, it avoids unjustly taxing workers on amounts they receive primarily on another’s behalf.
2. History of the Doctrine
The convenience of the employer doctrine was first recognized by administrative rulings in 1914—immediately after imposition of the federal income tax in 1913—in cases involving government employees who received in-kind lodging. But the doctrine quickly expanded to include private employees and cash housing allowances. In 1919, it was extended to in-kind lodging provided to private seamen. In 1920, it was extended in principle to all private employees. In 1921, it was extended by statute expressly to ministers. And in 1925—in the first federal court case addressing the doctrine—it was extended to cash housing allowances.
Early IRS rulings also extended the doctrine to cash allowances for volunteer charitable activities. In 1919, it was extended to a volunteer in the American Red Cross. And in the same year, it was extended to a clergyman under a vow of poverty. The non-economic motivation of these activities made it relatively easy to conclude that the allowances were primarily for the benefit of the general enterprise, not a private benefit to induce performance.
3. Codification in the Tax Code
In 1954, Congress codified some aspects the “convenience of the employer” doctrine in § 119(a)(2). Section 119(a)(2) now excludes the value of lodging from gross income for any employee—secular or religious—if five conditions are met. The lodging must be furnished 1) by an employer to an employee, 2) in kind, 3) on the business premises of the employer,
4) for the convenience of the employer, and 5) as a condition of employment. A wide variety of employees have qualified for this exemption, including construction workers, museum directors, an oil executive living in Tokyo, the president of the Junior Chamber of Commerce, a state governor, a rural school system superintendent, a prison warden, and many others.
But § 119(a)(2) is not the only provision codifying the convenience of the employer doctrine. Other provisions relax the requirements of § 119(a)(2) for certain types of employees. For example, § 119(c) governs “lodging in a camp located in a foreign country.” It defines “camp” in a way that eliminates the “business premises” and “condition of employment” factors. The rationale is that, when the camp is in a “remote area where satisfactory housing is not available on the open market,” the lodging is per se for the convenience of the employer.
Another per se rule applies to employees of educational institutions—such as college presidents, university faculty, or even elementary-school teachers. Under § 119(d), such employees can exclude a portion of the fair rental value of “qualified campus lodging,” even if they cannot satisfy any of the elements of the convenience of the employer doctrine. All they need to show is that the lodging is “(A) located on, or in the proximity of, a campus of the educational institution, and (B) furnished to the employee . . . by or on behalf of such institution for use as a residence.”
An even broader per se rule is § 134, which applies to members of the military. Under this provision, “any member or former member of the uniformed services” can receive tax-exempt housing benefits—including both in-kind lodging and cash allowances—regardless of whether the requirements of § 119(a)(2) are satisfied. This section codifies the reasoning in Jones that a service member’s duties “require his physical presence at his post or station; his service is continuous day and night; [and] his movements are governed by orders and commands.” Every service member is presumed to face these burdens on housing, whether living at home or abroad, on base or off, active duty or retired.
Nor is this per se rule limited to the military. Section 912 extends the same treatment to enumerated housing allowances of all government employees living abroad—including Peace Corps volunteers, CIA operatives, diplomats and consular officials, school teachers, and others. This reversed previous law, which required case-by-case application of the convenience of the employer doctrine to such employees.
Section 911 extends yet another per se rule to any “citizen or resident of the United States” residing in a foreign country. Such persons need not satisfy any of the requirements of § 119(a)(2); living abroad is enough. They can exclude housing costs above a certain level—whether housing is provided in-kind, through a cash allowance, or even purchased with their own funds. The basic rationale is that, if an individual is working abroad, she likely has significant extra housing costs that reduce her real income compared with a domestic worker. But a foreign worker need not prove that these considerations apply in her individual case.
Finally, under §§ 162 and 132, anyone posted away from her normal workplace for one year or less is not taxed on cash housing allowances or in-kind lodging provided by the employer. Again, there is no need to show that the lodging is used for work; the mere fact that she has moved away temporarily, while still maintaining her permanent home and primary business location, is enough to show that the temporary lodging is for the employer’s benefit.
In short, Congress has enacted a broad package of tax benefits designed to relieve workers who face unique, job-related housing requirements. The default rule is § 119(a)(2), which establishes a demanding, case-by-case test requiring all employees to demonstrate that their lodging is provided for the convenience of their employer. But Congress has also relaxed this default rule in a variety of situations where the type of work, the burdens on housing, or a non-commercial working relationship make it likely that the lodging was intended to benefit the employer.
4. Value of the Allowances
Critics of the parsonage allowance have suggested that these other exemptions apply only to a small number of secular groups. But according to congressional estimates, the annual value of these other exemptions vastly exceeds those under § 107. The projected value of other exemptions for 2017 totals more than $12 billion (including, for example, $6.4 billion for allowances for the armed forces and $2.3 billion for allowances for federal employees abroad), while the parsonage allowance is only expected to be worth about $800 million this year. The value of the parsonage allowance represents only a small fraction of the value of all exemptions for housing. All of these exemptions are reasonable reflections of the same overarching secular purpose of the convenience of the employer doctrine.
B. Ministers Fit Comfortably Within the “Convenience of the Employer” Doctrine
In light of this treatment of secular workers, the question under the plurality in Texas Monthly is simply whether it can be “fairly concluded that [ministers] could be thought to fall within the natural perimeter [of this legislation.]” The answer is a resounding “yes.”
A comparison to the strongest case—members of the military—is instructive. As Jones explained, a member of the military—whether living at home or abroad, on base or off, active duty or retired—is deemed to fall within the convenience of the employer doctrine on a per se basis because his duties “require his physical presence at his post or station; his service is continuous day and night; [and] his movements are governed by orders and commands.” Ministers face similar job-related demands on their housing.
1. Required Physical Presence
First, ministers are typically required to live at or near the church to be close to those they serve. This is most obvious in the case of Orthodox Jewish rabbis, who, due to Sabbath restrictions, must live within walking distance of the synagogue. It is also obvious in the case of religious orders, where leaders often live in the same convent or monastery as the members.
But it is also true in other settings. Many churches require their priests to live within the boundaries of the parish and near the church. Muslim imams usually must live near the mosque to lead prayer five times daily. Some churches are dedicated to serving a particular neighborhood, and the minister is expected to live in or near that neighborhood even when the location is undesirable. Still other churches assign ministers to serve in homeless shelters, hospitals, or nursing homes where they are expected to live in close proximity to those they serve. This sort of “voluntary displacement” has deep theological roots and, in the case of Christianity, is believed to mirror the incarnation of Christ.
On a more practical level, ministers in many small churches are the primary caretaker of the church building. Like the caretakers of apartment buildings—who often receive tax-free housing under § 119(a)(2)—ministers must respond when the fire alarm goes off, a pipe bursts, the furnace fails, the snow needs shovelling, or the building has other needs.
2. Service Day and Night
Ministers are also expected to be available to serve their congregations at any hour of the day or night. The Roman Catholic sacrament of anointing of the sick is administered only to those in danger of death. The sacrament must be administered “at the appropriate time,” and there are many “case[s] of necessity.” If the priest is not available at all hours, the sacrament cannot be administered. Ministers also respond at all hours to comfort grieving families, pray with congregants about emergencies, counsel spouses facing marital strife, hear confessions, and offer advice. The major life events of a congregation are not confined to regular business hours.
3. Use of Lodging for Their Duties
Ministers are also expected to use their homes to serve the church. In the Christian New Testament, there are two main lists of qualifications for ministers; both require them to be “hospitable.” In practice, this means hosting various church events, like Bible studies, women’s meetings, meals for new members, and the like. It also means providing temporary lodging for church members in transition, guest speakers, missionaries, and other travelers with a connection to the church—a practice frequently commended in the Christian New Testament. Many congregants also expect the minister’s home to be accessible for unplanned social visits.
Ministers also use their homes for church-related duties. When congregants seek comfort, prayer, counsel, confession, and advice—often at irregular hours—they often meet in the minister’s home. Counseling sessions, prayer meetings, and sensitive staff meetings are often held in the comfort of a home rather than a formal office. Meetings with lay leaders routinely occur in the home. Sermons are often prepared in the home. And in small churches that lack their own building, the only place to gather for worship is often the minister’s home.
4. Frequent Movement and Limited Choice
Ministers also face frequent movement and limited choice in their housing. This is most obvious in hierarchical denominations, such as Roman Catholic or Eastern Orthodox, where the placement of ministers is dictated by higher church authorities. In such churches, the diocesan Bishop often has absolute authority to move priests from parish to parish. Bishops can also agree to move priests across diocesan lines, including to foreign countries. Nor is frequent movement limited to hierarchical denominations. The average tenure of Southern Baptist ministers is less than 3 years, and for Mainline Protestant ministers it is only four years.
In many religious communities, the minister’s home is also expected to set an example of frugality. This is obviously true for members of religious orders who take a vow of poverty. But it also includes other religious groups, where a luxurious house may be viewed as a sin or a distraction from the minister’s pastoral service. In other cases, ministers may be obliged to live in an area with housing costs far higher than the minister would otherwise choose. Either way, the housing costs are driven by the needs of the church, not the personal consumption choices of the minister.
The point of describing ministers’ activities is not to show that ministers are exactly like military service members in every respect. It is that they are in a unique, non-commercial employment relationship with unique, job-related demands on their housing. Given this reality, Congress could “fairly conclude that [ministers] could be thought to fall within the natural perimeter” of the convenience of the employer doctrine. Accordingly, § 107(2) is constitutional even under the more stringent test of the Texas Monthly plurality.
In that sense, challenges to the parsonage allowance are analogous to Rojas v. Fitch. There, the First Circuit considered an Establishment Clause challenge to religious exemptions from federal and state unemployment taxes. The plaintiff argued that these exemptions provided unique, unjustified benefits to religious employers in violation of Texas Monthly. The First Circuit, however, disagreed. Applying the Texas Monthly plurality, it held that a religious tax exemption is permissible as long as similar exemptions are “conferred upon a wide array of non-sectarian groups . . . in pursuit of some legitimate secular end.” Turning to the unemployment taxes at issue, the court noted that the federal and state insurance programs “exclud[e] from coverage a variety of workers whose employment patterns are irregular or whose wages are not easily accountable.” Although the plaintiff argued that these exemptions were underinclusive and thus effectively favored religion, the court rejected the argument that “a provision incidentally benefitting religion must grant a like benefit to every group that could also conceivably fall within the secular rationale for the exemption provision.” Rather, it was enough that the exemptions “serve the legitimate secular purpose of facilitating the administration of the unemployment insurance system” and reduce “entanglement concerns.”
Here, the fit between § 107 and the “legitimate secular purpose” of the convenience of the employer doctrine is even stronger. The exemption has a far longer historical pedigree, and the value of the exemption is dwarfed by the value of a wide array of nonreligious exemptions. If the exemption considered in Rojas was legal under Texas Monthly—as it surely was—then the parsonage allowance has an even greater claim to legitimacy.
C. To the Extent that the Parsonage Allowance Provides Special Treatment to Ministers, It Is Justified by Important First Amendment Principles
Not only do ministers fit comfortably within the convenience of the employer doctrine, but there are powerful reasons for addressing the taxation of ministers separately in § 107, and not simply lumping them in with all other employees under § 119. Indeed, just as Congress adapted the convenience of the employer doctrine to employees in foreign camps (§ 119(c)), educational institutions (§ 119(d)), military service (§ 134), overseas government jobs (§ 912), overseas private jobs (§ 911), and jobs requiring temporary displacement (§§ 162 and 132), it has adapted the doctrine to ministers—and it has very good secular reasons for doing so. Specifically, § 107 serves two critical secular purposes: reducing entanglement between church and state, and avoiding discrimination among religious groups. Both purposes are not just constitutionally permissible but laudable.
1. The Tax Code Routinely Provides Special Treatment to Churches and Ministers To Reduce Entanglement and Discrimination Among Religions
Objections to the parsonage allowance implicitly assume that churches and ministers are in an ordinary employment relationship, and so any tax provision addressing them separately is automatically suspect. But that assumption is flawed. In many cases, the First Amendment not only permits “special solicitude” for churches, but requires it. In particular, the First Amendment 1) restricts government interference in the relationship between churches and ministers, 2) forbids government entanglement in religious questions, and 3) prohibits government discrimination among denominations. These three values—church autonomy, non-entanglement, and non-discrimination—are reflected throughout the tax code in specific protections for churches, none of which are available to secular non-profits.
For example, several provisions protect the relationship between churches and ministers by exempting churches from paying or withholding certain types of taxes:
- Churches are not required to withhold federal income taxes from ministers in the exercise of ministry.
- Churches are exempt from Social Security and Medicare taxes for wages paid to ministers in the exercise of ministry; instead, ministers are uniformly treated as self-employed.
- Churches are exempt from state unemployment insurance funds authorized by the Federal Unemployment Tax Act.
Other provisions protect church autonomy by exempting churches from disclosing information:
- Churches and certain related entities are not required to file Form 990, which discloses sensitive financial information.
Still others reduce entanglement by offering unique procedural protections:
Still others modify tax provisions so that they apply neutrally among various church polities:
- Churches can maintain a single church benefits plan exempt from ERISA for employees of multiple church affiliates, regardless of common control, and for ministers, regardless of their employment status. This is designed “[t]o accommodate the differences in beliefs, structures, and practices among our religious denominations.”
- Churches can include ministers in 403(b) contracts (a type of tax-deferred retirement benefit), even if ministers do not qualify as employees.
- Churches can provide certain insurance to entities with common religious bonds, even if those entities are not structured to meet normal common control tests.
Congress has been particularly careful to make sure that general tax rules do not discriminate among ministers based on the nature of their relationship with the church. For example, when Congress extended eligibility for social security to ministers in 1954, it stipulated that all ministers would be treated as self-employed, regardless of whether they were common-law employees—precisely to avoid discriminating between groups based on the status of their ministers as employees.
In short, the tax code does not treat churches and ministers as ordinary employers and employees. Rather, Congress has crafted numerous tax provisions that apply only to churches and ministers. These provisions, like § 107(2), reduce entanglement and prevent discrimination among religions.
2. The Parsonage Allowance Reduces Entanglement
Some critics of the parsonage allowance claim that § 107 might increase entanglement because it requires the government to inquire into religious doctrine to determine who is a minister. But this is mistaken. Viewed in context of the entire tax code, the per se parsonage allowance of § 107 is far less entangling than the next best alternative—applying the notoriously difficult multi-factor standard of § 119 to ministers.
Whenever the government taxes churches and ministers, there is no completely disentangling alternative: “Either course, taxation of churches or exemption, occasions some degree of involvement with religion.” To figure out which alternative is best, it is essential to distinguish between two types of entanglement. One is called “enforcement entanglement.” It occurs when the government taxes churches, and is therefore required to value church property, place liens on church property, and (in some cases) foreclose on church property. This creates direct confrontations between church and state and threatens church autonomy.
The other type of entanglement is called “borderline” entanglement. It occurs when the government exempts churches, and is therefore required to decide who qualifies for the exemption and who doesn’t. For example, it may have to decide whether an entity is “religious” and whether a publication is “consistent with ‘the teaching of the faith.’” Policing the borders of a complicated exemption threatens to entangle courts in religious questions.
These two types of entanglement are illustrated by Walz and Texas Monthly. Walz focused on “enforcement entanglement.” There, the Court explained that taxing churches “would tend to expand the involvement of government by giving rise to tax valuation of church property, tax liens, tax foreclosures, and the direct confrontations and conflicts that follow in the train of those legal processes.” Exempting churches, by contrast, would “restrict the fiscal relationship between church and state,” thus “tend[ing] to complement and reinforce the desired separation insulating each from the other.”
Texas Monthly focused on “borderline entanglement.” There, all periodicals were subject to tax, except those that consisted “wholly of writings promulgating the teaching of [a] faith.” Because the government had to decide which messages were “consistent with ‘the teaching of the faith,’” the exemption produced “greater state entanglement” than providing no exemption at all.
Section 107 reduces both enforcement entanglement and borderline entanglement. It reduces enforcement entanglement because it avoids the “latent dangers inherent in the imposition of . . . taxes” on a core part of the relationship between churches and their ministers. More importantly, it reduces borderline entanglement because it replaces the notoriously fact-intensive standard of § 119 with the bright-line rule of § 107.
Section 119 is extremely difficult, if not impossible, to apply to ministers. First, it requires the minister to qualify as an “employee” under IRS rules. This, in turn, requires the government to tax differentially depending on internal matters of church polity. If the minister belongs to a denomination that gives him broad autonomy or exposes him to significant economic risk, he may fail this test and be considered self-employed. Some decisions suggest that United Methodist Council ministers would qualify as employees, but Assemblies of God and various Pentecostal ministers would not. Even if a minister qualified as an employee, a § 119 exemption would be unavailable if one entity provided the housing (such as the congregation), but a different entity qualified as the “employer” (such as the diocese)—thus pressuring churches to make decisions about church structure based on definitions in the tax code.
Once these threshold concerns are overcome, § 119 still requires the government to decide whether a minister’s housing was “furnished for the convenience of the employer” as “a condition of his employment.” This, in turn, requires the government to decide whether the lodging is truly necessary “to enable him properly to perform the duties of his employment.” In other words, is it really necessary for a priest or imam “to be available for duty at all times”? Is it really necessary for a rabbi to live in close proximity to the synagogue, to counsel synagogue members at home, to host meetings at home, and to prepare derashot at home? These sorts of inquiries are extremely difficult and fact-intensive for secular employees. They raise grave constitutional concerns when applied by the government to evaluate the relationship between a church and its ministers.
Section 107, by contrast, recognizes that the government cannot decide which uses of a minister’s home are “necessary” to the mission of the church and which are not. It asks only whether the employee is functioning as a minister. This is an inquiry courts have been conducting for decades—not only in the tax context, but also under the First Amendment “ministerial exception.” Indeed, it is an inquiry that the Supreme Court itself said was constitutionally required just five years ago.
This is why § 107 is easily distinguishable from the exemption in Texas Monthly. There, the alternative to the religious exemption for periodicals was no exemption at all—all periodicals would be taxed equally. Thus, striking down the religious exemption eliminated any possibility of borderline entanglement. By contrast, if § 107 were struck down, the alternative would be to apply § 119 to ministers. Far from eliminating borderline entanglement, that would exacerbate it.
3. The Parsonage Allowance Reduces Discrimination
Section 107(2) also reduces discrimination among religions. The Supreme Court has repeatedly held that nondiscrimination is “[t]he clearest command of the Establishment Clause.” This applies not just to intentional discrimination among religions, but also to “indirect way[s] of preferring one religion over another.” Of course, a facially neutral law is not invalid merely because it has a greater “incidental effect” on one denomination than another. But “when the state passes laws that facially regulate religious issues”—as § 107 clearly does—“it must treat individual religions and religious institutions without discrimination or preference.”
The leading case on religious nondiscrimination is Larson. There, a Minnesota law imposed reporting requirements on all charitable organizations, but it exempted “religious organizations that received more than half of their total contributions from members.” This had the effect of distinguishing between “well-established churches,” which received ample “financial support from their members,” and “churches which are new and lacking in a constituency” and had to rely on “public solicitation.” The state defended its rule on the ground that it was “based upon secular criteria” and merely “happen[ed] to have a disparate impact upon different religious organizations.” But the Supreme Court rejected this argument, concluding that the statute “focuses precisely and solely upon religious organizations” and makes “explicit and deliberate distinctions between [them].”
Maintaining the parsonage allowance without applying it to cash allowances for clergy housing—§ 107(1) without § 107(2)—would have the same effect. “[W]ell-established churches” with “financial support” can afford to purchase a parsonage and provide tax-free housing to ministers. But “churches which are new and lacking in a constituency” cannot. This creates a serious disparity between wealthy and poor denominations.
But the disparity is not merely financial. The decision to have a parsonage is also influenced by theological considerations. In some denominations, like the Roman Catholic Church, the use of church-owned parsonages is fundamental to how the Church deploys its clergy. In other denominations—typically newer and less hierarchical ones—there is no historical or theological emphasis on church-owned parsonages. Sometimes, this is because churches expect ministers to be bi-vocational; other times, it is because churches may take years before they establish a permanent place of worship; still other times, it is because the churches have a theological reluctance to amass large holdings of worldly property. And in some cases, ministers are expected to be itinerant, making a housing allowance the only feasible way of meeting their housing needs. Given these differences among denominations, limiting § 107 only to in-kind housing only discriminates along theological, not just financial, lines.
Thus, it is no surprise that equal treatment of housing allowances was first imposed by courts to avoid religious discrimination, even before Congress enacted § 107(2). This occurred in the early 1950s, when three federal courts held that cash housing allowances must be excluded from the income of ministers. Congress then codified these decisions in § 107(2). When it did so, it expressly stated that it was seeking to “remove the discrimination in existing law” among various denominations, as had been required in the federal court decision.
Treating cash allowances and in-kind housing equally is also logical. Although cash payments may be compensatory, they need not be. “[J]ust as an employee is often furnished tangible property which cannot be regarded as compensation, an employee may be furnished cash which is not compensation.” The question is whether the lodging is furnished for the convenience of the employer, not whether it is cash or in-kind. Thus, it is no surprise that the first court decision involving the convenience of the employer doctrine rejected a distinction between cash allowances and in-kind housing. So did the first court of appeals decision involving ministers. So did early IRS rulings on charitable volunteers. And so did early commentators. Indeed, § 119 is the only housing allowance to distinguish between cash and in-kind housing benefits. There is no reason to import this distinction into § 107, especially when it causes discrimination among religions.
V. Striking Down the Parsonage Allowance Would Endanger Scores of Tax Provisions Throughout Federal and State Law
An interpretation of the Establishment Clause that invalidates the parsonage allowance also threatens numerous other provisions throughout federal and state tax codes. As discussed above, nearly every state in the nation provides some tax exemptions for religious groups without analogous exemptions for other nonprofit institutions. Likewise, Congress has created a host of tax provisions that treat churches and ministers differently than other employers and employees in order to protect the First Amendment values of church autonomy, non-entanglement, and non-discrimination.
To take just one example, the federal tax code includes a religious exemption from self-employment taxes for “a duly ordained, commissioned, or licensed minister of a church” who “is conscientiously opposed to, or because of religious principles . . . is opposed to, the acceptance . . . of any public insurance.” This statutory test—“duly ordained, commissioned, or licensed”—is identical to the Treasury Regulation definition of a “minister of the gospel” for purposes of § 107. Thus, if § 107 impermissibly advances religion or entangles the government in religious questions, then so does the self-employment tax exemption for religious objectors to Social Security. But “[w]ithout this exemption in the Code, the IRS would be required to enforce the self-employment tax against individuals despite their religious opposition to ‘public insurance’ such as the Social Security system financed by the self-employment tax.”
Surely the First Amendment requires no such thing. Indeed, multiple courts have rejected this argument. Yet this is the clear implication of the position that § 107 violates the Establishment Clause. Critics of the parsonage allowance are wrong. The Establishment Clause does not require such hostility to religion. These numerous state and federal tax provisions are constitutional, as is the parsonage allowance.
Suggested citation: Hannah C. Smith & Daniel Benson, When a Pastor's House Is a Church Home: Why the Parsonage Allowance Is Desirable Under the Establishment Clause, 18 Fed. Soc'y Rev. 60 (2017).
 Revenue Act of 1921, Pub. L. No. 67-98, § 213(b)(11), 42 Stat. 227, 239 (1921).
 26 U.S.C. § 107.
 It is also called the “parsonage exemption” or “parsonage exclusion.”
 See, e.g., Freedom From Religion Foundation v. Geithner, 644 F.3d 836 (9th Cir. 2011); Freedom From Religion Foundation v. Lew, 773 F.3d 815 (7th Cir. 2014); Gaylor v. Lew, No. 3:16-cv-215-bbc (W.D. Wis. Apr. 5, 2016); Erwin Chemerinsky, The Parsonage Exemption Violates the Establishment Clause & Should Be Declared Unconstitutional, 24 Whittier L. Rev. 707 (2003).
 U.S. Const. amend. I, cl. 1.
 Lamb’s Chapel v. Ctr. Moriches Union Free Sch. Dist., 508 U.S. 384, 395 (1993) (describing Lemon test); Lynch v. Donnelly, 465 U.S. 668, 688-89 (1984) (O’Connor, J. concurring) (first articulating “no endorsement” test).
 See, e.g., Doe v. Elmbrook Sch. Dist., 687 F.3d 840, 869-77 (7th Cir. 2012) (Easterbook & Posner, JJ., dissenting from en banc decision) (calling Lemon and “no endorsement” test “hopelessly open-ended”); Gerard V. Bradley, Protecting Religious Liberty, 42 DePaul L. Rev. 253, 261 (1992); Richard W. Garnett, Religion, Division, & the First Amendment, 94 Geo. L.J. 1667 (2006); Douglas Laycock, Towards a General Theory of the Religion Clauses, 81 Colum. L. Rev. 1373, 1380-88 (1981); Michael W. McConnell, Religious Freedom at a Crossroads, 59 U. Chi. L. Rev. 115, 118-20 (1992).
 McCreary Cty. v. Am. Civil Liberties Union of Ky., 545 U.S. 844 (2005).
 See, e.g., Mount Soledad Mem’l Ass’n v. Trunk, 132 S. Ct. 2535, 2535-36 (2012) (Alito, J., dissenting from denial of certiorari) (“Establishment Clause jurisprudence is undoubtedly in need of clarity.”); Utah Highway Patrol Ass’n v. Am. Atheists, Inc., 132 S. Ct. 12, 14-15, 17, 19, 21-22 & n.3 (2011) (Thomas, J., dissenting from denial of certiorari) (“Establishment Clause jurisprudence [is] in shambles,” “nebulous,” “erratic,” “no principled basis,” “purgatory,” “impenetrable,” “ad hoc patchwork,” “limbo,” “incapable of consistent application,” “our mess,” “little more than intuition and a tape measure.”) (citations omitted); Lamb’s Chapel, 508 U.S. at 398 (Scalia, J., concurring); Allegheny Cty. v. Am. Civil Liberties Union, 492 U.S. 573, 655-57 (1989) (Kennedy, J., concurring in judgment in part and dissenting in part); Corp. of Presiding Bishop of Church of Jesus Christ of Latter-day Saints v. Amos, 483 U.S. 327 (1987) (O’Connor, J., concurring in judgment); Wallace v. Jaffree, 472 U.S. 38, 107-13 (1985) (Rehnquist, J., dissenting); id. at 90-91 (White, J., dissenting); Comm. for Public Ed. & Religious Liberty v. Regan, 444 U.S. 646, 671 (1980) (Stevens, J., dissenting).
 Van Orden v. Perry, 545 U.S. 677, 686 (2005) (plurality); see also Town of Greece v. Galloway, 134 S. Ct. 1811 (2014) (not applying Lemon); Zelman v. Simmons-Harris, 536 U.S. 639 (2002) (same); Good News Club v. Milford Cent. Sch., 533 U.S. 98 (2001) (same).
 Town of Greece, 134 S. Ct. at 1819; Hosanna-Tabor Evangelical Lutheran Church & Sch. v. E.E.O.C., 565 U.S. 171, 183 (2012); Van Orden, 545 U.S. at 686; Marsh v. Chambers, 463 U.S. 783, 786 (1983).
 397 U.S. 664 (1970).
 489 U.S. 1 (1989).
 Walz, 397 U.S. at 675-680; Texas Monthly, 489 U.S. at 11-13.
 Town of Greece, 134 S. Ct. at 1819 (emphasis added).
 Id. at 1823, 1825, 1819.
 463 U.S. 783, 786 (1983).
 Id. at 790.
 565 U.S. 171, 183-84 (2012).
 545 U.S. 677, 686 (2004) (plurality opinion); see also id. at 699-700 (Breyer, J., concurring) (looking to “national traditions” and the monument’s historical context).
 397 U.S. at 680.
 Id. at 675.
 Id. at 680.
 Id. at 681.
 Id. at 682, 685, 687.
 See Michael McConnell, Establishment and Disestablishment at the Founding, Part 1: Establishment of Religion, 44 Wm. & Mary L. Rev. 2105, 2131 (2003); see also Felix v. City of Bloomfield, 847 F.3d 1214 (10th Cir. 2017) (Kelly, J., and Tymkovich, C.J., dissenting from denial of rehearing en banc) (relying on Professor McConnell’s historical analysis).
 McConnell, supra note 26 at 2146-59.
 397 U.S. at 675.
 Id. at 676.
 2 Va. Stat. at Large of 1792-1806 (Shepherd) 149.
 9 Va. Stat. at Large (1775-78, Hening) 351; 13 Va. Stat. at Large (1789-92, Hening) 112, 241, 336-37; 2 Va. Stat. at Large of 1792-1806 (Shepherd) 149.
 Walz, 397 U.S. at 683 (Brennan, J., concurring).
 Acts of the Corporation of the City of Washington, First Council, c. V, approved Oct. 6, 1802, at 13.
 Walz, 397 U.S. at 676.
 6 Stat. 116 (1813); 6 Stat. 600 (1834).
 6 Stat. 162 (1816).
 6 Stat. 675 (1836).
 Act of June 17, 1870, 16 Stat. 153.
 Texas Monthly, 489 U.S. at 31 (Scalia, J., dissenting).
 Haw. Rev. Stat. § 244D-4(b)(4); S.D. Codified Laws § 35-5-6(2); Wash. Rev. Code § 66.20.020(3).
 Cal. Rev. & Tax. Code Ann. § 6363.5; Idaho Code § 63-3622J; Md. Ann. Code, Tax-Gen. § 11-206(d)(1)(ii).
 Mich. Comp. Laws § 205.54a(b)(ii); Mo. Rev. Stat. § 144.450(4); Va. Code § 58.1-3617.
 See Texas Monthly, 489 U.S. at 31 n.3 (Scalia, J., dissenting) (collecting statutes).
 Hobbie v. Unemployment Appeals Comm’n, 480 U.S. 136, 144-45 (1987).
 Corp. of the Presiding Bishop, 483 U.S. at 335.
 397 U.S. at 692 (Brennan, J., concurring).
 Id. at 674.
 Id. at 673.
 Id. at 676.
 489 U.S. at 5.
 Id. at 15 (plurality).
 Id. at 11.
 Id. at 25-26 (White, J., concurring).
 Id. at 28.
 Id. at 27-28.
 Id. at 28 (Blackmun, J., concurring).
 430 U.S. 188, 193 (1977). Accord Freedom From Religion Foundation v. Lew, 983 F. Supp. 2d 1051, 1061-62 (W.D Wisc. 2013) (Blackmun/O’Connor concurrence is controlling); Catholic Health Initiatives Colorado v. City of Pueblo, Dep’t of Fin., 207 P.3d 812, 828 (Colo. 2009) (Eid, J., dissenting) (same).
 See infra Part III.
 26 U.S.C. § 119(a).
 § 119(c).
 § 119(d).
 § 134.
 § 912.
 § 911.
 §§ 162, 132.
 Texas Monthly, 489 U.S. at 28, 26 (emphasis added) (citing Follett v. McCormick, 321 U.S. 573 (1944) and Murdock v. Pennsylvania, 319 U.S. 105 (1943)).
 Id. at 26, 28.
 See id. at 26 (“spreading the gospel”); id. (“spread the gospel”); id. (“publications”); id. (“religious literature”); id. at 27 (“religious books”); id. (“religious books”); id. (“religious literature”); id. (“philosophical literature”); id. at 28 (“taxation of books and journals”); id. (“religious literature”); id. (“religious literature”); id. (“religious publications”); id. (“religious messages”); id. (“dissemination of religious ideas”); id. at 29 (“religious literature”); id. (“religious literature”); id. (“atheistic literature”); id. (“religious literature”).
 489 U.S. at 14 n.4 (emphasis added).
 Id. at 17 (emphasis added) (quoting Walz, 397 U. S. at 696).
 See Lemon v. Kurtzman, 403 U.S. 602, 612 (1971) (statute (1) “must have a secular legislative purpose,” (2) “its principal or primary effect must be one that neither advances nor inhibits religion,” and (3) it “must not foster an excessive government entanglement with religion.”).
 United States v. Gotcher, 401 F.2d 118, 121 (5th Cir. 1968) (emphasis added).
 See Treas. Reg. § 1.132-5(a)(1)(v); Treas. Reg. § 1.162-2(a)–(b).
 Jones v. United States, 60 Ct. Cl. 552, 575 (1925); see generally J. Patrick McDavitt, Dissection of a Malignancy: The Convenience of the Employer Doctrine, 44 Notre Dame L. Rev. 1104 (1969).
 Id. at 1105 (citing T.D. 2079, 16 Treas. Dec. Int. Rev. 249 (1914)).
 Id. at 1106 (citing O.D. 265, 1 C.B. 71 (1919)).
 Id. (citing Treas. Reg. 45, art. 33 (1920 ed.); T.D. 2992, 2 C.B. 76 (1920)).
 Revenue Act of 1921, Pub. L. No. 98, § 213(b)(11), 42 Stat. 227, 239 (overturning O.D. 862, 4 C.B. 85 (1921)).
 Jones, 60 Ct. Cl. 552.
 O.D. 11, 1919-1 C.B. 66.
 O.D. 119, 1919-1 C.B. 82.
 Treas. Reg. § 1.119-1(b).
 Treas. Reg. § 1.119-1(f) Ex. (7); Stone v. Comm’r, 32 T.C. 1021 (1959).
 Jane Zhao, Nights on the Museum: Should Free Housing Provided to Museum Directors Also Be Tax-Free?, 62 Syracuse L. Rev. 427, 447-49 (2012).
 Adams v. United States, 218 Ct. Cl. 322 (1978).
 U.S. Jr. Chamber of Commerce v. United States, 167 Ct. Cl. 392 (1964).
 Rev. Rul. 75-540, 1975-2 C.B. 53; See also Rev. Rul. 90-64, 1990-2 C.B. 35 (principal representative of the U.S. to a foreign country).
 Haack v. United States, 75-2 U.S.T.C. ¶ 9847 (S.D. Iowa 1975).
 I.R.S. Priv. Ltr. Rul. 9126063 (March 29, 1991).
 Compare 26 U.S.C. § 119(c) with § 119(a)(2).
 § 119(c)(2)(A).
 § 119(d)(2)–(3).
 § 134.
 60 Ct. Cl. at 569.
 McDavitt, supra note 78 at 1108 & n.40 (collecting decisions).
 See Staff of the Joint Committee on Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2015-2019 (Comm. Print 2015) at Table 1. The value of temporary-location costs under §§ 162 and 132 is unknown; it appears to be reported within the larger category of “fringe benefits,” totaling $7.5 billion. Id. Allowances for Armed Forces and federal employees include more than just housing.
 489 U.S. at 17.
 60 Ct. Cl. at 569.
 Henri J.M. Nouwen et al., Compassion: A Reflection on the Christian Life 60-73 (2005).
 1983 Code c.1004 § 5.
 Code c.1001.
 Code c.999, 1000 § 1, 1003 § 3.
 Titus 1:8; 1 Timothy 3:2 (Revised Standard Version).
 See, e.g., Matthew 10:11 (lodging for apostles); Acts 16:15 (lodging for missionaries); Romans 16:2 (lodging for Phoebe); 3 John 1:5-8 (lodging for traveling Christians).
 The Barna Group, Report Examines the State of Mainline Protestant Churches (Dec. 7, 2009), https://www.barna.com/research/report-examines-the-state-of-mainline-protestant-churches/.
 See Alison Smale, Vatican Suspends German Bishop Accused of Lavish Spending on Himself, N.Y. Times, Oct. 23, 2013.
 Texas Monthly, 489 U.S. at 17 (plurality).
 127 F.3d 184 (1st Cir. 1997), abrogated on other grounds by Hardemon v. City of Boston, No. 97-2010, 1998 WL 148382 (1st Cir. Apr. 6, 1998).
 Id. at 188 (quoting Texas Monthly, 489 U.S. at 14-15).
 Id. at 188.
 Id. at 189.
 Hosanna-Tabor, 565 U.S. at 189.
 Walz, 397 U.S. at 674.
 Larson v. Valente, 456 U.S. 228, 244 (1982).
 26 U.S.C. § 3401(a)(9).
 26 U.S.C. §§ 1402(c)(4), 1402(e), 3121(b)(8).
 26 U.S.C. § 3309(b)(1).
 26 U.S.C. § 6033(a)(3).
 26 U.S.C. § 7611.
 26 U.S.C. § 508(a), (c)(1)(A).
 26 U.S.C. § 414(e).
 See 26 U.S.C. § 414(e)(3)(B), (5)(A); Miscellaneous pension bills: Hearings before the Subcommittee on Private Pension Plans and Employee Fringe Benefits of the Committee on Finance, United States Senate, Ninety-Sixth Congress, First Session (Dec. 4, 1979), at 367 (Statement of Sen. Talmadge).
 26 U.S.C. § 403(b)(1)(A)(iii).
 26 U.S.C. § 501(m)(3)(C)-(D); G.C.M. 39874 (May 4, 1992); Treas. Reg. § 1.502-1(b).
 Conf. Rep. No. 83-2679 (1954).
 Walz, 397 U.S. at 674.
 Edward A. Zelinsky, Do Religious Tax Exemptions Entangle in Violation of the Establishment Clause?, 33 Cardozo L. Rev. 1633 (2012).
 Id. at 1640.
 Id. at 1635.
 Texas Monthly, 489 U.S at 20.
 Walz, 397 U.S. at 698-99 (Harlan, J., concurring).
 Id. at 674.
 Id. at 676.
 489 U.S. at 5.
 Id. at 20 (plurality).
 See Walz, 397 U.S. at 673.
 See Weber v. Comm’r, 103 T.C. 378 (1994), aff’d, 60 F.3d 1104 (4th Cir. 1995); Shelley v. Comm’r, T.C.M. (RIA) 1994-432 (1994); Alford v. United States, 116 F.3d 334 (8th Cir. 1997).
 See Fuhrmann v. Comm’r, T.C.M. 1977-416 (1977).
 Treas. Reg. § 1.119-1(b).
 McDavitt, supra note 78 at 1139-40.
 Hosanna-Tabor, 565 U.S. at 190 (prohibiting “government interference with internal church decisions that affect the faith and mission of the church itself”); id. at 206 (Alito, J., concurring) (courts cannot asses “the importance and priority of the religious doctrine in question,” what a “church really believes,” or “how important that belief is to the church’s overall mission”); Corp. of Presiding Bishop, 483 U.S. at 342 (Brennan, J., concurring) (courts cannot “determin[e] that certain activities are in furtherance of an organization’s religious mission”).
 See McClure v. Salvation Army, 460 F.2d 553 (5th Cir. 1972).
 Hosanna-Tabor, 565 U.S. 171.
 It is no answer to say that § 119 applies only to in-kind lodging. Cash allowances present the same entanglement problem under §§ 162 and 280A(c)(1).
 Larson, 456 U.S. at 244, 246 (collecting cases).
 Fowler v. Rhode Island, 345 U.S. 67, 70 (1953).
 See Employment Div. v. Smith, 494 U.S. 872, 878 (1990).
 Colo. Christian Univ. v. Weaver, 534 F.3d 1245, 1257 (10th Cir. 2008) (internal quotations marks omitted).
 456 U.S. at 231.
 Id. at 246 n.23.
 Id. (internal quotations marks omitted).
 MacColl v. United States, 91 F. Supp. 721, 722 (N.D. Ill. 1950); Conning v. Busey, 127 F. Supp. 958 (S.D. Ohio 1954) (following MacColl); Williamson v. Comm’r, 224 F.2d 377 (8th Cir. 1955).
 H.R. Rep. No. 83-1337, at 4040 (1954); S. Rep. No. 83-1622, at 4646 (1954). Congress did the same thing with housing allowances for government workers living overseas to eliminate discrimination among them. In the 1950s, many overseas employees received tax-exempt, in-kind housing, but some did not. Congress enacted the Overseas Differential and Allowances Act authorizing cash housing allowances, and § 912 excluding those cash housing allowances from income. Thus, § 912 does the same thing for overseas employees that § 107(2) does for ministers. See Anderson v. United States, 16 Cl. Ct. 530, 534 (1989), aff’d, 929 F.2d 648 (Fed. Cir. 1991). See also id. at 535 (“Congress intended that all federal overseas employees be treated uniformly.”).
 Williamson, 224 F.2d at 379 (quoting Saunders v. Comm’r, 215 F.2d 768, 771 (3d Cir. 1954)).
 Jones, 60 Ct. Cl. 552.
 Williamson, 224 F.2d at 379.
 O.D. 11, 1919-1 C.B. 66; O.D. 119, 1919-1 C.B. 82.
 See McDavitt, supra note 78 at 1132-33, 1138 (distinction is “artificial and formalistic” and has “no practical place in the convenience of the employer doctrine”).
 Parts I-III, supra.
 26 U.S.C. § 1402(e).
 See Treas. Reg. § 1.107-1(a) (citing Treas. Reg. § 1.1402(c) 5.
 Zelinsky, supra note 134 at 1669.
 See, e.g., Droz v. Comm’r, 48 F.3d 1120, 1124-25 (9th Cir. 1995) (upholding § 1402(g) against an Establishment Clause challenge); Ballinger v. Comm’r, 728 F.2d 1287, 1292-93 (10th Cir. 1984) (rejecting Establishment Clause challenge to § 1402(e)).