Supporters visiting Javier Flores, an undocumented immigrant from Mexico living in sanctuary at Arch Street United Methodist Church, Philadelphia

Bastiaan Slabbers/NurPhoto/Getty Images

Supporters visiting Javier Flores (right), an undocumented immigrant from Mexico living in sanctuary at Arch Street United Methodist Church, Philadelphia, March 2017

Midway through last year, half a dozen Latin American immigrants scattered across the United States received “notices of intent to fine” from US Immigration and Customs Enforcement (ICE) for failing to leave the country. ICE has long had the authority to levy financial penalties for immigration violations, but these were extraordinary: first, because they were enormous—up to $799 for each day after a final order of deportation was issued, and running as high as $500,000 per person—and second, because all those singled out were living in churches. Although the number of such sanctuary residents in the United States has not exceeded four dozen at any one time since this ancient practice was revived by immigration activists, it is a highly visible form of defiance that the executive branch clearly finds vexing. (The fines were withdrawn four months later without explanation, then partially reinstated in December.) The very roundaboutness of the confrontation was noteworthy: even under the Trump administration, no US sanctuary church has been directly subjected to any enforcement action.

Ecclesial sanctuary was used as a protection from punishment throughout Christendom for more than a thousand years before entering a period of prohibition and decline beginning in the sixteenth century. The expanding jurisdiction of early-modern states, as well as the Reformation and the Enlightenment, made an institution dependent on the prerogatives of clerical power increasingly unsustainable. Statesmen, historians, and the Catholic Church’s canon lawyers had largely come to condemn sanctuary as a “hoary privilege” that might once have been needed to mitigate cruel punishments amid a “barbarous state of society” but had turned into “[an invitation] to crime” and a “public nuisance.”

The Vatican’s Codex Juris Canonici (Code of Canon Law), which was promulgated in 1917 and superseded eight hundred years of accumulated church legislation, still defended the extinct practice: “A church enjoys the right of asylum, and those who flee there may not be extracted, other than in the case of necessity, without the permission of the [bishop], or else the rector of the church.” When the Codex was updated in 1983, even this limp vestige—lacking threats of excommunication, ceding “necessity” to the discretion of secular authority—vanished, just as sanctuary was experiencing a sort of rebirth. Refugees fleeing civil wars in Guatemala and El Salvador had come to the attention of religious-minded US activists who began to provide them help in crossing over from Mexico, navigating the immigration bureaucracy, and, sometimes, finding temporary residence in church buildings.

In March 1982 these activists, primarily progressive Catholics and Protestants, publicly established the Sanctuary Movement, which became a nonhierarchical network of hundreds of congregations across the country. As the Central American asylum crisis waned, the movement quietly dissipated, then reformed a decade ago and underwent dramatic growth in response to the surges in deportations under Presidents Obama and Trump. The Sanctuary Movement had inspired similar activism among churches in Western and Northern Europe in the 1980s that were responding to a wave of asylum-seekers from the developing world; this activism has likewise been revived in recent years by the Mediterranean refugee crisis.

The prologue to the Sanctuary Movement of the 1980s was a limited effort among Vietnam War–era churches to provide refuge to draft evaders and AWOL servicemembers, whom military police did not hesitate to drag out. This was expected; such sanctuary was understood purely as moral confrontation, proclaiming a righteous and sheltering higher law. Although there is undoubtedly a romance in the idea of drawing a bright line between what is God’s and what is Caesar’s, church sanctuary practices should not be seen, according to the sociologist Randy Lippert, as part of “a majestic and eternal conflict between two monolithic entities—church…and state.” What allows sanctuary to function today is a tangled and contingent array of legal, political, cultural, and administrative factors. This is, as recent scholarship reveals, as it ever was.

Greek asylums, often connected to a particular divinity and stifled under Roman rule, were the antecedent of the asylum tradition that developed in the Christianized empire. Established and state-regulated by the late fourth century, Christian sanctuary was based in episcopal intercession and penitential discipline; it was intended to spare the body the worst consequences of crime and thereby to save the soul from the everlasting implications of sin. In addition to criminals avoiding punishment, debtors and slaves made frequent use of sanctuary; masters could reclaim their slaves by swearing not to retaliate against them, and in extreme situations the church would purchase them or arrange for their resale.

During the Anglo-Saxon period, sanctuary served primarily as a buffer to settle blood feuds. Thereafter, more or less from the Assize of Northampton in 1176, which established the public prosecution of crime and linked sanctuary to the use of judicial exile, through the onset of the Black Death in 1348, which precipitated a drop in executions for felony, English sanctuary’s conventions consolidated around royal law, and its numbers held steady. A rough estimate suggests there were a thousand English sanctuary-takers a year.1


A felon’s flight to sanctuary was usually a matter of recent offense and hot pursuit. Once on holy ground, he had forty days to submit to trial, to arrange for the king’s pardon, or to accept permanent exile. (The forty-day rule likely originated in Anglo-Saxon practice, to allow time to negotiate blood money, though it could also facilitate reflection and penitence.) Those found guilty of a felony were hanged, so avoiding trial was mainly the point. Pardons were easier for the well connected, and the felons were mostly of modest or poor means, so abjuring England was overwhelmingly the outcome.

The road to deportation began with the summoning of the king’s coroner, an official charged with recording pleas, who took possession of the felon’s property and administered the oath of abjuration: to depart the realm and not to return except by leave of the king or his heirs. Girt in sackcloth, shoeless, and holding a small rude cross, the abjuror proceeded to a port prescribed by the coroner, with a small allowance for food and lodging, inviolable but with the “wolf’s head” of outlawry and summary execution hanging over him if he were to stray from the highway—not a picturesque pilgrim’s journey but an actively monitored process. On reaching his destination, the abjuror was supposed to wade into the sea and cry out for passage (the main route in this high period of abjuration was from Dover to Wissant, just across the Channel).2

Eligible sanctuaries ranged from field churches to cathedrals. The boundaries within which a fugitive was exempt from seizure, and the fines for violating them, were generally standardized by the category of church. But some sites possessed exceptional protections, marked by crosses as far as a mile and a half from the high altar, with penalties increasing incrementally; apprehending a fugitive who gained the frithstool (the “chair of peace” beside the altar) was held to be an unemendable offense, for which the pursuer’s life was, in theory, forfeit. The interpretation of boundaries could produce elaborate drama: in 1579 a fugitive debtor near Seville was assigned the role of Christ in his sanctuary’s mystery play and assured he would remain safe while on the church’s cart in the square, but he was pushed off by Judas, who had apparently been bribed by the creditor. (A judge ordered the debt forgiven.)

Although English sanctuary welcomed an exceptionally broad range of felons, from traitors to cheaters at dice, from herring thieves to “a boye that kyld a byge boye…with the horling of a stone,” technical restrictions applied to seekers. Sanctuary was a popular destination for jailbreakers, but those already convicted were ineligible, a fact unhappily learned by men who, having been inadequately hanged, revived after the gallows and fled to a church. Conversely, lack of sufficient criminality was a bar: one common-law commentator held that

certain timid men who seek the protection of the Church, although they have not committed a felony…will be well advised to acknowledge that they have slain someone whom they will be able to produce afterwards alive and well.

In 1270 a chaplain, while having sex with another man’s wife, grew fearful that he was being watched and fled to a church, where he was told he could not abjure for fornication; having now implicated himself and fearing the cuckold’s retaliation, he confessed to having stolen, from a chest in which he had briefly hidden, the felony-threshold sum of eightpence.

Augustine, a great proponent of sanctuary, proposed that “the good do not flee the good, nor the just the just; the unjust flee the just, the just flee the unjust, or the unjust flee the unjust.” This early idea of sanctuary as welcoming equally the culpable and the legally blameless (sinners, all, at any rate) had a rhetorical vigor that long outlived reality; the English historian Henry Hallam, writing in 1818, argued that sanctuary had been “as often a shield to innocence, as an impunity to crime” before it was a decadent institution for the guilty. The press today, responding to the predicament of migrants, continues to portray traditional sanctuary as a resource for the unjustly persecuted. But it had little to do with “innocence” for most of its mature history.

One problem in both popular understanding and early historical scholarship is that “timid men” who may have quietly availed themselves of the temporary safety of the church, secured by general prohibitions against sacrilege, have been conflated with felons and the rare nonfelon like our chaplain invoking the right of sanctuary. Another such problem was that certain elites, having attracted the crown’s disfavor, took refuge against peremptory seizure. In Thomas More’s History of King Richard III, the Duke of Buckingham, coaxing the queen consort to send the second of Richard’s nephews out of Westminster to his doom, deliberately confuses these two different forms of sanctuary: “He can be no Sanctuary-man that had neither wisdome to desire it, nor malice to deserue it, whose life or libertie, can by no lawfull processe stand in jeopardie.”


It was not, of course, lawful process that presented the risk. In felony sanctuary, the fugitive sought relief from the neutral application of the king’s law, and the sovereign was in no way a maleficent figure—it was he who piously sanctioned the protection; but his whim was another matter. The Wars of the Roses witnessed the life-saving resort to nonfelony asylum by leaders on both sides of the conflict, and such events likely introduced the resilient trope of sanctuary as a defense against tyranny.

Political sanctuary-takers with powerful connections in the clergy usually turned up at the great minsters and abbeys of England, known as chartered sanctuaries after the royal grants—supposedly dating from deep in the Anglo-Saxon period—that imbued them with special jurisdictional liberties. These sanctuaries are a particular focus of the historian Shannon McSheffrey’s Seeking Sanctuary: Crime, Mercy, and Politics in English Courts, 1400–1550 (2017), which is devoted to English sanctuary’s late period. By the early fourteenth century, the development of formal debt litigation created a class of fugitives who sought out the chartered sanctuaries and enjoyed a new type of protection from seizure while resident: unlimited in duration. Within the century, this privilege, with some clerical fudging of history, was extended to felons.

Though the sleight of hand succeeded, the conceit that criminals might slip in and out of these sanctuaries to commit mischief quickly became a line of attack by civic authorities; the London liberties of Westminster and St. Martin’s-le-Grand were said to be an intolerable menace. (At the latter, sanctuary men felt free to drink at an adjoining tavern as long as they kept one hand on the abbey wall.) Indeed, the most florid denunciations of sanctuary, contemporary and retrospective, seem to have been directed at chartered sanctuaries; though forty-day sanctuary was thereby implicated, it is difficult to imagine that abjuration invited such invective about avoiding just punishment and inducing crime.3 Exile was effective at removing offenders, and a hard mercy.

McSheffrey finds that sanctuary recovered its strength after the mid-fifteenth century and was thriving just prior to its downfall, its numbers stable or rising in most of the early Tudor period. Sanctuary had even found a new use: facilitating pardons for aristocrats whose internecine affrays now fell under the purview of felony law. McSheffrey insists that hostility to sanctuary was far from universal, and just as previous kings who fought other clerical privileges had tolerated it, Henry VIII was not, according to the preamble of the 1540 statute “Concerning Sanctuaries,” seeking “the utter abolishing and extinguishment of all sanctuaries and the privileges of the same,” as these had proved “very expedient and convenient.”

Sanctuary had for centuries grown detached from penitential discipline and become incorporated within royal criminal law, while the church’s emphasis shifted from the sacrality of space to jurisdictional immunity, which in turn became more closely linked to royal grants. By its end, felony sanctuary was overseen from registration to deportation by lay managers, and even that archenemy of sanctuary Thomas Cromwell, Henry’s chief minister, occasionally limited his aim to expunging solely the religious element—a suggestion that also appears in a tract attributed to Luther. Henry’s feud with Rome did not require abolitionist fervor over a legal convention to which the church was ever more tenuously connected.

A series of new laws, culminating under Henry but beginning decades earlier, did restrict sanctuary. They declared ineligible many seekers on the basis of exceptions variously devised by the church’s canonists on the Continent or preserved from Roman code by scholars of civil law, and went further by canceling protections for many serious felonies. Nocturnal destroyers of fields, ravishers of virgins, purple dye fish collectors, subtle pickers of locks, horse thieves, and sacriligists were now excluded. The civilian commentators also began pushing a line the canonists had explored centuries earlier: punishment, which was undermined by sanctuary, could have a preventive effect on crime.

Yet in McSheffrey’s account, new restrictions were not what killed off sanctuary in England. The decisive blow was the monastic dissolutions of 1540, which destroyed the chartered sanctuaries and the substantial independent jurisdictions they represented, combined with a reform of abjuration: foreign exile was replaced with a paltry number of slots in eight unhappily designated English cities, and abjurers were now branded with an “A” on the right thumb. Parliamentary abolition of felony sanctuary followed in 1603 and again, emphatically, in 1624, while sanctuary for debt straggled on for another two centuries. In certain Continental jurisdictions, more often Catholic, forms of criminal sanctuary survived until the late eighteenth and early nineteenth centuries. In the Spanish New World, the plea of Iglesia me llamo (“My name is Church,” invoking the right to be returned to sanctuary) was recorded as late as the eve of World War I.

How long English sanctuary would have continued to thrive amid the centralization of the Tudor monarchy, and with similar trends underway elsewhere, may be an unproductive question. McSheffrey’s more interesting suggestion about sanctuary’s fate considers an adjacent institution: benefit of clergy, which was originally a plea for removal to ecclesiastical jurisdiction, became a commutation of punishment based on a literacy test available to nonclergy, then a de facto memorization test (except for despised defendants, the Latin passage was always Psalm 51:1), and finally, after it was extended to women, a general lenience for first-time offenders. With a little more time, could sanctuary, too, have been wholly absorbed by the state?

Modern sanctuary has added a curious chapter to this history. The Sanctuary Movement of the 1980s began with a handful of congregations near the Mexican border bailing out detained Central American refugees and assisting them with asylum claims. It soon turned into a clandestine cross-border transport operation; the advocates rightly observed that even when the authorities did not thwart the application process outright, grants of asylum were perilously rare for those fleeing US client regimes. (Under one percent of Guatemalans were approved in 1984, compared with 61 percent of Iranians and about a fifth of all nationalities.) With the publicity around these activities and the movement’s rapid growth, some three thousand sites offered temporary housing to refugees, from Native American reservations to Trappist monasteries. Today’s sanctuary cities, whose protection consists in disregard of immigration status and refusal to assist with federal enforcement, are an enduring outgrowth of this era.

A few thousand (out of a million-odd) Central American refugees were aided directly in crossing borders and through resettlement, and many others through indirect means, including charitable work, political pressure, and litigation: the movement’s singular victory came from a lawsuit by church groups that finally pushed the INS, in 1990, to reevaluate Guatemalan and Salvadoran asylum applications. But sanctuary had been in decline and riven by disputes since another major court case, in which border sanctuary workers who had been infiltrated by undercover federal agents were charged with illegally transporting aliens. This group of clergy and church lay leaders advanced free-exercise arguments about the religious obligation to provide shelter and aid, and they asserted that they had been securing the rights the refugees were entitled to under US statutes and international treaties. They were convicted in 1986 but served no prison time. After the movement dissipated, the Illegal Immigration Reform and Immigrant Responsibility Bill of 1996 resulted in an estimated one million family separations in the subsequent decade.

The current sanctuary movement formed amid that legacy and under the threat of further regressive legislation. After modest beginnings in 2006, it swelled during the deportation push of Obama’s second term to encompass a number of churches that approached 1980s levels, and that has more than doubled since 2016, to a coalition of about 1,100 congregations. They meet asylum-seekers at the border to offer assistance, accompany out-of-status residents to ICE check-ins, and promote know-your-rights campaigns. ICE itself has routinely sought them out as caretakers for “catch and release” asylum claimants awaiting adjudication. The movement’s members, as in the 1980s, skew white and liberal, although (as was also the case then) there is significant participation by minority churches and non-Christian faiths.4 Politically conservative congregations of the sort occasionally seen in the 1980s movement are no longer evident, owing partly to an enthusiasm for deportation that has consumed the right.

Reflecting the migration demographics of recent decades, much of the current movement’s larger ministry has been oriented toward long-term migrants without legal status and with family members who are US citizens. The same is true of physical church sanctuary. In the 1980s the “underground railroad” of migrant assistance, not sanctuary itself, was the flagship project; the movement’s invocations of abolitionism, the Nuremberg principles (to disobey unlawful orders), and biblical hospitality had broad cultural currency yet reflected the particular situation: vulnerable fugitives escaping cruel authority. The seekers were helpless in an unfamiliar country, and when they actually stayed in churches, sometimes for days or weeks or only as a convenient place to sleep, it typically facilitated long-term resettlement (the movement’s demand of the government was “extended voluntary departure,” i.e., permission to remain until it was safe to go home). Those few in what was called “public sanctuary” were exemplary victims of US foreign policy’s emboldened client regimes and testified to host congregations and media about death squads whose violence achieved a vivid excess of terror and humiliation.

Today’s sanctuary-takers tend to be upstanding Americans in all but paperwork, and to speak of jobs disrupted and fears of family separation. Virtually all of them would have been eligible for leniency under a new policy of prosecutorial discretion late in the Obama administration, and some have become activist leaders in their own communities. Some have applied for asylum, but most have not. For them, sanctuary is not an arrival; it is the final defense before they are forced to abjure their adoptive homeland. In 2017 and 2018, between a quarter and half of US sanctuarians received a stay of deportation or other relief from some branch of government, but those numbers have lately fallen.

Western and Northern European churches have since the 1980s used sanctuary in precisely this way: as a last stand against deportation, aimed at obtaining a reprieve (usually for applicants denied asylum). Defensive sanctuary of this sort is condoned by the authorities in Germany, making it more mundane and more popular than in the United States, while in other instances it can seem almost magical. Dutch police are legally proscribed from disrupting organized worship, so last year a church in The Hague defeated the deportation of an Armenian family by holding a single continual service for three months.

In the United States, the undergirding of protection is fairly diffuse. ICE’s 2011 Sensitive-Locations Memo discourages enforcement actions at churches, schools, and hospitals, following INS precedent dating to 1981, when agents outraged a Los Angeles congregation by chasing an undocumented man inside the church and arresting him in the loft. Today the avoidance of such “sacrilege” is partly about avoiding censure from the one major Christian demographic that has never participated in sanctuary: white evangelicals. But another factor, rarely mentioned, is that ICE has no mechanism for obtaining judicial warrants against violators of civil immigration codes and, theoretically, cannot enter private property to enforce its administrative warrants without permission or improbable exigency. This is true of private homes as well, but agents are less likely to employ underhanded tactics when churches are involved.5 Anyone for whom an arrest warrant can be obtained—i.e., someone wanted for a crime—would be, in a historical reversal, an implausible candidate for sanctuary.

The past decade has seen a trend toward “crimmigration”—the criminalizing of reentry violations and aggressive use of the criminal justice system to police and restrict immigration6—and a related preoccupation with immigrants’ dangerousness. Another historical reversal, arising from this, is that most people in public sanctuary are women, who are more likely than men to evoke sympathy and less likely to become the coddled criminals of deportationist rallying cries. The 2015 death of a San Francisco woman named Kate Steinle after she was hit by a ricocheted bullet fired by a man with seven felonies and five deportations (who was ultimately acquitted of killing her) became a cudgel against sanctuary cities. The death last summer of a Colorado man whose motorcycle was struck by a former church sanctuary resident driving on a suspended license was touted with particular glee because the victim had vocally opposed the Trump administration’s immigration policies. And the first sanctuary-seeker of the Obama era, Daniel Neyoy Ruiz, took refuge twice at Tucson’s Southside Presbyterian Church while awaiting stays of deportation, then vanished from public view; a search of the Arizona courts shows his subsequent years marked by several misdemeanor convictions.

In sheltering a deportable person, congregations could find themselves exposed to federal harboring charges, though intimidation through fines—which the churches, too, could receive—is far less arduous than criminal litigation. Churches might fall back on the First Amendment free-exercise clause and its expanded protections created by the 1993 Religious Freedom Restoration Act (RFRA) against “substantial burden” on worship. The defense against harboring would depend, in most federal circuits, on the openness with which sanctuary was undertaken, and whereas the RFRA was not available to defendants in the 1985–1986 trial, the sanctuary movement now avoids engaging in such active, visible violations of federal law. However, there are currently about the same number of people in undeclared, clandestine sanctuary as there are in open sanctuary.

University of Toronto

‘Seeking Sanctuary’; lithograph after a painting by Ralph Hedley, 1890

Many in the 1980s movement framed law-breaking actions less as civil disobedience than as “civil initiative,” aimed at the observance of good laws the Reagan administration was ignoring—mainly the 1951 Convention Relating to the Status of Refugees and the 1980 Refugee Act—whereby aliens were entitled to asylum because of their “well-founded fear of persecution.” The civil initiative advocates went further, arguing for the innate legitimacy of the Central Americans’ claims by rejecting the possibility that they might be “economic migrants.” One of the founders of the movement, who helped conduct interviews in Mexico to establish refugees’ credibility, called it “an emergency alternative to the INS.”

Church deference to sovereign decrees about asylum eligibility is an echo of history, and not just superficially. If sanctuary left no legacy in criminal process, it did leave one in these conventions regulating the person of the asylum-seeker. (The word “refugee” originated in the English crown’s grants of asylum to Huguenots, an act that mirrored the church’s prerogative of mercy.) Refugee laws, then, are where traditional sanctuary was absorbed by the state, and a self-conscious “sanctuary” movement arose to negotiate with it. The perversity of this genealogy is thoroughly traced in the legal scholar Simon Behrman’s Law and Asylum (2018). Behrman contends that the Sanctuary Movement’s misstep was to see refugee law as a fully benevolent protection—tinted by the heroism that church sanctuary itself sometimes enjoys—whereas functionally, in providing an escape from one sovereign, it inevitably subjects the petitioner to the whim of a second.

In the current movement, moral legitimacy appears less closely linked to law. This may be not so much a philosophical position as a practical one: our system for evaluating refugees, and therefore differentiating them from mere immigrants or economic migrants, becomes ever more byzantine, ad hoc, and prone to change. Whereas in the 1980s most applications and appeals turned on complaints of state oppression, today they have to do far more often with gang violence and domestic abuse precedents. These were shoehorned into the asylum schema a decade ago, then overruled two years ago by the attorney general, but are still largely in use, if with generally poor odds and highly inconsistent outcomes. And the immigration system may yet be challenged by vast, morally persuasive claims that cannot be finessed into the US asylum structure—for example, those of climate refugees.

Notwithstanding this less traditional relationship to power, US church sanctuary has lately returned to a defensive form more recognizable from history. And if there is a specific precedent for spending three years or more in sanctuary, as several already have, it may be the chartered sanctuary. McSheffrey speculates that had sanctuary continued to evolve naturally, it might have ended up as a sentence of incarceration. Certainly lengthy sanctuary stays have been described as jail-like, by fugitives both in the Tudor era and today. Gainful employment in chartered sanctuary was difficult, and accommodations were not free; the average resident may have lasted six months.

Today’s sanctuary-takers are typically supported by congregational fund-raising, but the long-term challenges of sanctuary have discouraged new seekers. Last fall a Peruvian woman in extended sanctuary in Colorado delivered a baby in the church’s rec room; a Mexican woman who recently marked 1,200 days in sanctuary, also in Colorado, stayed there throughout the chemotherapy, death, and funeral of her mother, who had returned to Mexico; and a Mexican woman and her three teenage children endured serious cases of Covid in a Philadelphia church this spring. Many, say organizers, have chosen to go underground and hope for a new administration. Better than to live what the poet Michael Drayton in 1603 called “a poore dead life” where “hate stands without.”

The Kingdoms of Israel and Judah contained several Cities of Refuge, which were permitted to take in men who had killed by accident, who might later return home from exile following the death of the high priest in Jerusalem. This convention had no influence on either early Christian asylums or classical sanctuary, though it has become a touchstone for the recent movements in their defense of apparently blameless outlaws.

Modern “sanctuary” is a thoroughly synthetic anachronism—defined neither by sacred space, on which the practice was originally predicated and which developed in Christianity with the emergence of saints’ cults and relics, nor by the church’s intercession for the sinful and the guilty, mediated by clerical and tutelary power. It is defined by the way it draws strength from the church membership. It is a profoundly participatory phenomenon. And today congregants routinely say that it is they who have been saved and succored by the work of protecting strangers. “Less a means to shield a man, more a means to expose a church…to make a church really a church,” as the Reverend William Sloane Coffin Jr., whose congregations provided refuge during both the Vietnam War and the Central American crisis, once put it. Sanctuary now does something Augustine never envisioned: it allows the church to be the guilty party.