Martial Law in the United States: Its Meaning, Its History, and Why the President Can’t Declare It

On August 20, 1942, milit­ary police in Honolulu, Hawaii, arres­ted a man named Harry White. Under normal circum­stances, the U.S. milit­ary would not have been involved in his case. He was a stock­broker, not a soldier, and neither he nor his busi­ness had any connec­tion with the armed forces. Even his alleged crime — embez­zle­ment of funds from a client — was a viol­a­tion of civil­ian, not milit­ary, law.foot­note1_7uas­fnz1

But noth­ing about Hawaii was normal in 1942. It had been under martial law since the Japan­ese attack on Pearl Harbor in Decem­ber 1941.foot­note2_bf6o­gen2  Its courts were closed and replaced with milit­ary tribunals. The rules govern­ing every­day life were set not by an elec­ted legis­lature but by the milit­ary governor. The army controlled every aspect of life in the islands, from crim­inal justice to park­ing zones and curb­side trash removal.foot­note3_0p7c­qyf3

White was brought before a milit­ary prov­ost court. His attor­ney objec­ted to the court’s juris­dic­tion, reques­ted a jury trial, and asked for time to prepare a defense. But Major Murrell, the presid­ing milit­ary officer, rejec­ted these motions. Instead, just five days after being arres­ted, White was tried without a jury, convicted, and sentenced to five years in prison.foot­note4_wqx4qpj4

As White’s story illus­trates, martial law — a term that gener­ally refers to the displace­ment of civil­ian author­it­ies by the milit­ary — can be and has been employed in the United States. Indeed, federal and state offi­cials have declared martial law at least 68 times over the course of U.S. history.foot­note5_joz5h6q5  Yet the concept has never been well under­stood. The Consti­tu­tion does not mention martial law, and no act of Congress defines it. The Supreme Court has addressed it on only a hand­ful of occa­sions, and the Court’s reas­on­ing in these decisions is incon­sist­ent and vague.foot­note6_fd05ocm6  The preced­ents are also old: the most recent one — in which the Court over­turned Harry White’s convic­tion — was decided almost 75 years ago.

This report aims to clear up the confu­sion that surrounds martial law. To do so, it draws on recent legal schol­ar­ship, the few rules that can be gleaned from Supreme Court preced­ent, and general prin­ciples of consti­tu­tional law. It concludes that under current law, the pres­id­ent lacks any author­ity to declare martial law. Congress might be able to author­ize a pres­id­en­tial declar­a­tion of martial law, but this has not been conclus­ively decided. State offi­cials do have the power to declare martial law, but their actions under the declar­a­tion must abide by the U.S. Consti­tu­tion and are subject to review in federal court.

Outside of these general prin­ciples, there are many ques­tions that simply cannot be answered given the sparse and confus­ing legal preced­ent. Moreover, although lack­ing author­ity to replace civil­ian author­it­ies with federal troops, the pres­id­ent has ample author­ity under current law to deploy troops to assist civil­ian law enforce­ment. Until Congress and state legis­latures enact stricter and better-defined limits, the exact scope of martial law will remain unsettled, and the pres­id­ent’s abil­ity to order domestic troop deploy­ments short of martial law will be danger­ously broad.

“Martial law” has no estab­lished defin­i­tion.foot­note1_mfja7zw7

In the United States, however, the milit­ary’s domestic activ­it­ies typic­ally fall into one of three categor­ies. First, the armed forces some­times assist civil­ian author­it­ies with “non–law enforce­ment” func­tions. For example, in the after­math of Hurricane Katrina in 2005, the milit­ary deployed heli­copters along the Gulf Coast to carry out search-and-rescue missions that local govern­ments were unable to do them­selves. Second, and far less frequently, the milit­ary assists civil­ian author­it­ies with “law enforce­ment” activ­it­ies. For example, state and federal troops were deployed to help police suppress the 1992 Los Angeles riots. Third, on some occa­sions, the milit­ary has taken the place of the civil­ian govern­ment. This is what happened in Hawaii during World War II.

Usually, but not always, the term “martial law” refers to the third category. It describes a power that, in an emer­gency, allows the milit­ary to push aside civil­ian author­it­ies and exer­cise juris­dic­tion over the popu­la­tion of a partic­u­lar area. Laws are enforced by soldiers rather than local police. Policy decisions are made by milit­ary officers rather than elec­ted offi­cials. People accused of crimes are brought before milit­ary tribunals rather than ordin­ary civil­ian courts. In short, the milit­ary is in charge.foot­note2_zl9l­bxg8

This is a dramatic depar­ture from normal prac­tice in the United States. The U.S. milit­ary, when allowed to act domest­ic­ally at all, is ordin­ar­ily limited to assist­ing civil­ian author­it­ies. Martial law turns that rela­tion­ship on its head. The displace­ment of civil­ian govern­ment distin­guishes it from other emer­gency powers, such as the suspen­sion of the writ of habeas corpus. Suspend­ing the writ allows the govern­ment to detain and hold indi­vidu­als without charge but does not imply any unusual role for the armed forces. While a declar­a­tion of martial law might be accom­pan­ied by a suspen­sion of habeas corpus, they are distinct concepts.

Martial law has not always meant what it does today. The term first appeared in England in the 1530s during the reign of King Henry VIII.foot­note3_bfm37m99  At that time and for centur­ies after­ward, martial law gener­ally referred to what is now called “milit­ary law.”foot­note4_3cdh­deg10  This is the law that applies when a soldier is court-martialed. In the modern United States, it is codi­fied in the Uniform Code of Milit­ary Justice.foot­note5_nkn2ea711

U.S. law did not recog­nize martial law as an emer­gency power until the mid-19th century. Before that time, the idea of allow­ing milit­ary rule in an emer­gency was considered outrageous — as evid­enced by the national reac­tion to the first declar­a­tion of martial law in U.S. history. In Decem­ber 1814, toward the end of the War of 1812, Gen. Andrew Jack­son led a small army in the defense of New Orleans against a much larger invad­ing Brit­ish force. As part of his defens­ive prepar­a­tions, Jack­son imposed martial law on the city. He censored the press, enforced a curfew, and detained numer­ous civil­ians without charge. Moreover, he contin­ued milit­ary rule for more than two months after his famous victory at the Battle of New Orleans had ended any real threat from the Brit­ish.foot­note6_0yfkjz812

Jack­son argued that his actions were justi­fied because the govern­ment in New Orleans had ceased to func­tion as a result of the impend­ing Brit­ish attack, leav­ing the milit­ary as the only body able to protect the city. In that situ­ation, he claimed, the milit­ary had the author­ity to do anything that was “neces­sary” to preserve New Orleans.foot­note7_aet0ij213  This was a novel argu­ment, and it did little to explain why he kept the city under martial law for so long.

At the time, almost every­one rejec­ted Jack­son’s theory, which perhaps is unsur­pris­ing. The found­ing gener­a­tion had been deeply suspi­cious of milit­ary power. That suspi­cion is appar­ent in the Declar­a­tion of Inde­pend­ence, which accuses King George III of render­ing “the Milit­ary inde­pend­ent of and super­ior to the Civil power” — and in the Consti­tu­tion, which poin­tedly divides the war powers between Congress and the pres­id­ent, and requires that the commander in chief always be a civil­ian.foot­note8_6qa0utf14

In an 1815 case, the Louisi­ana Supreme Court described Jack­son’s conduct in New Orleans as “tramp­ling upon the Consti­tu­tion and laws of our coun­try.”foot­note9_9p05kys15  Simil­arly, acting Secret­ary of War Alex­an­der Dallas explained in a letter to Jack­son that martial law had no legal exist­ence in the United States outside of the Articles of War, the prede­cessor to the modern Uniform Code of Milit­ary Justice.foot­note10_ciqc25516  Over­all, the consensus in 1815 was that martial law was simply another term for milit­ary law, and that milit­ary juris­dic­tion could extend no further than the armed forces them­selves.

After Jack­son relin­quished control of New Orleans back to its civil­ian govern­ment, the local federal district judge held him in contempt of court, fining him $1,000. Jack­son paid the fine, and for the next 27 years, noth­ing more came of the incid­ent. However, in the early 1840s, the now-aging former pres­id­ent orches­trated a campaign in Congress to refund him the cost of the fine, plus interest.foot­note11_3f4s­jp117

The ensu­ing congres­sional refund debates marked the begin­ning of a shift in how Amer­ic­ans under­stood martial law. By pursu­ing a refund, Jack­son hoped to set a preced­ent for, as one histor­ian put it, “the legit­im­acy of viol­at­ing the Consti­tu­tion and civil liber­ties in times of national emer­gency.”foot­note12_fd6g1r618  He got exactly what he wanted. Congress enacted the refund bill in Febru­ary 1844, symbol­ic­ally endors­ing Jack­son’s three-month-long impos­i­tion of martial law in New Orleans almost 30 years after it had ended.foot­note13_kg89bix19

By this time, the United States’ second exper­i­ence with martial law was already under­way in Rhode Island. The so-called “Dorr War” involved a dispute over the state’s first consti­tu­tion, which severely restric­ted the right to vote. In 1842, after efforts to reform this system had been rebuffed for years, a large group of Rhode Islanders led by Thomas Dorr organ­ized its own consti­tu­tional conven­tion, adop­ted a new consti­tu­tion, held elec­tions, and declared itself the true govern­ment of Rhode Island. When Dorr rallied his support­ers to assert their author­ity by force, the Rhode Island General Assembly declared martial law and called out the state mili­tia to suppress the rebel­lion.foot­note14_7mp0o6u20

In 1849, the U.S. Supreme Court upheld the legal­ity of Rhode Island’s martial law declar­a­tion in Luther v. Borden.foot­note15_jt84inf21

Writ­ing for the major­ity, Chief Justice Roger Taney — of Dred Scott infamy — embraced Andrew Jack­son’s idea that martial law allows civil­ians to be subjec­ted to milit­ary juris­dic­tion in an emer­gency. He described this power as an essen­tial part of states’ right to defend them­selves and sugges­ted that it is inher­ent to all sover­eign govern­ments.foot­note16_b41ic6e22  By endors­ing the consti­tu­tion­al­ity of martial law, the Supreme Court finished what Congress had star­ted with the refund bill. The Luther decision makes clear that martial law exists as an emer­gency power that can be invoked in the United States, at least by state legis­latures.foot­note17_18qmx8123

But Luther also leaves many ques­tions unanswered. It does not explain the legal basis for martial law, its scope, when it may be declared, or who is author­ized to declare it. Indeed, the Supreme Court has never directly held, in Luther or any subsequent case, that the federal govern­ment has the power to impose martial law. In one case, the Court sugges­ted in “dicta”— a term for language in a judi­cial opin­ion that is not a neces­sary part of the hold­ing and is not strictly legally bind­ing — that the federal govern­ment may declare martial law.foot­note18_0y6qrno24  It assumed the same in another case, but only for the purpose of decid­ing a narrower legal ques­tion.foot­note19_98el­g0w25  Neither of those decisions conclus­ively affirms that a federal martial law power exists.

Indeed, the Supreme Court has never directly held, in Luther or any subsequent case, that the federal govern­ment has the power to impose martial law.

Over time, however, consist­ency of prac­tice has papered over gaps in the legal theory. The United States made extens­ive use of martial law during the Civil War, impos­ing it on border states like Missouri and Kentucky where U.S. forces battled with Confed­er­ate insur­gents.foot­note20_ukfet8w26  The Confed­er­acy, too, relied on it heav­ily.foot­note21_n4n44yd27  The prac­tice did not end with the war: in the 90 years between the start of the Civil War and the end of World War II, martial law was declared at least 60 times.foot­note22_tga7sol28  What had been mani­festly uncon­sti­tu­tional in the eyes of the Louisi­ana Supreme Court in 1815 had become a relat­ively ordin­ary part of Amer­ican life by the end of the 19th century.

States — and state governors in partic­u­lar — have declared martial law far more often than the federal govern­ment. In the 1930s, Oklahoma Governor William “Alfalfa Bill” Murray declared martial law at least 6 and perhaps more than 30 times during his tenure.foot­note23_jdeigqz29  City mayors and gener­als within states’ National Guard forces have also declared martial law on occa­sion. However, no state legis­lature has done so since the Rhode Island General Assembly in 1842.

Not all of the milit­ary deploy­ments under these declar­a­tions included what we today consider the defin­ing feature of “martial law” — the displace­ment of civil­ian author­ity. Many cases involved the use of the milit­ary to rein­force local police. In other cases, however, troops effect­ively replaced the police, and in some instances, they were used to impose the will of state or local offi­cials rather than to enforce the law.

State offi­cials have some­times declared martial law in response to viol­ent civil unrest or natural disasters, such as the Akron Riot of 1900 or the 1900 Galve­ston hurricane.foot­note24_f80t38m30  Far more often, however, they have used martial law to break labor strikes on behalf of busi­ness interests. For example, in Septem­ber 1903, at the request of mine owners, Color­ado Governor James Peabody declared martial law in Cripple Creek and Tellur­ide to break a peace­ful strike by the West­ern Feder­a­tion of Miners. The Color­ado National Guard conduc­ted mass arrests of strik­ing work­ers and detained them in open-air bull pens. The Guard even ignored state court orders to release the pris­on­ers, with one officer declar­ing, “To hell with the consti­tu­tion.”foot­note25_ukhro9531

States’ use of martial law contin­ued well into the 20th century, reach­ing a peak in the 1930s — a decade that also saw an increase in the flag­rant abuse of this power by governors. In 1933, for example, Geor­gia Governor Eugene Talmadge declared martial law “in and around” the headquar­ters build­ing of the state High­way Board as part of a scheme to force out some of the board’s commis­sion­ers, whom he had no legal power to remove. This “coup de high­way depart­ment” was ulti­mately success­ful. Remark­ably, Talmadge’s successor, Governor Eurith Rivers, tried to do the same thing in 1939, but his attempt failed.foot­note26_jtxm3gi32

Misuses of martial law were not confined to Geor­gia. In 1931, Texas Governor Ross Ster­ling engaged in a stan­doff with the federal courts over his govern­ment’s abil­ity to enforce a regu­la­tion limit­ing oil produc­tion by private well oper­at­ors. At the climax of the conflict, Ster­ling imposed martial law on several counties — despite the total absence of viol­ence or threats of viol­ence — and deployed the Texas National Guard to enforce the regu­la­tion. He declared that the federal courts had no power to review his decision. The U.S. Supreme Court disagreed, explain­ing that “[t]here is no such avenue of escape from the para­mount author­ity of the Federal Consti­tu­tion.”foot­note27_iz6r4ee33  It ordered Texas to stop using the milit­ary or any other means to enforce the regu­la­tion.

The federal govern­ment has used martial law far less frequently than the states, impos­ing it only a few times since the end of Recon­struc­tion. Gener­als have declared it more often than the pres­id­ent, such as in 1920, when U.S. Army Gen. Fran­cis C. Marshall imposed martial law on Lexing­ton, Kentucky, in order to suppress a lynch mob attempt­ing to storm the court­house.foot­note28_2k6n­lja34  Most recently, the federal govern­ment declared martial law in Hawaii after the Japan­ese attack on Pearl Harbor in 1941, which initi­ated three years of abso­lute milit­ary rule in the islands.foot­note29_umscqt335

As abruptly as it took hold in the mid-19^th century, martial law disap­peared from Amer­ican life after World War II. The federal govern­ment has not declared martial law since it restored civil­ian rule to Hawaii in 1944. At the state level, martial law was last declared in 1963, when Mary­land Governor J. Millard Tawes imposed it on the city of Cambridge for more than a year in response to clashes between racial justice advoc­ates and segreg­a­tion­ists.foot­note30_7qabe3736  But even if the power to declare martial law has not been used in decades, it still exists in the case law and in the record books — and it remains poorly under­stood.

Despite the wide­spread use of martial law in the century that followed the Supreme Court’s ruling in Luther, many of the legal ques­tions surround­ing it remain unanswered. The Court has never explained the legal basis for martial law. It has implied that the federal govern­ment can declare it but has never said so conclus­ively. When discuss­ing the possib­il­ity of a federal martial law power, the Court has never clearly indic­ated whether the pres­id­ent could unilat­er­ally declare martial law or if Congress would first need to author­ize it.

Inso­far as the Supreme Court has said anything on these ques­tions, its state­ments have been incon­sist­ent. During the 19th century, the Court sugges­ted in dicta that a federal martial law power was “implied in sover­eignty” or justi­fied by “neces­sity.”foot­note1_jlj070437  In the early 20th century, it seemed to believe that the state martial law power was related to the exec­ut­ive’s consti­tu­tional power to “execute the law.”foot­note2_737y­urn38  During World War II, the Court assumed (without decid­ing) that Congress could author­ize a federal declar­a­tion of martial law but did not make clear whether that author­iz­a­tion was required.foot­note3_j6slcio39  In contrast, in a much earlier but influ­en­tial 1866 concur­ring opin­ion, Chief Justice Salmon Chase did conclude that federal martial law exists and that it must be author­ized by Congress.foot­note4_mcb8odc40

Look­ing only at the Supreme Court’s martial law decisions, one can pick and choose from them to argue for almost anything. The case law is incon­sist­ent and too sparse for a clear pattern in the Court’s reas­on­ing to emerge. It is also old: even the most recent Supreme Court decision on martial law — Duncan v. Kahanamoku, decided in 1946 — pred­ates many signi­fic­ant devel­op­ments in U.S. consti­tu­tional law.foot­note5_rshsyk341  The preced­ents thus provide little help in determ­in­ing the legal basis for martial law — or, assum­ing that federal martial law is even permiss­ible, whether its use is controlled by Congress or the pres­id­ent.

One must there­fore turn to a differ­ent Supreme Court case for guid­ance: Young­stown Sheet & Tube Co. v. Sawyer.foot­note6_sdfx2gy42  Decided in 1952, Young­stown held that Pres­id­ent Harry Truman could not seize U.S. steel mills to prevent a labor dispute from disrupt­ing the nation’s supply of steel in wartime. The decision — and Justice Robert Jack­son’s concur­ring opin­ion in partic­u­lar — has since become the doctrinal lode­star for assess­ing any exer­cise of power by the exec­ut­ive.foot­note7_jq9tkkg43  Consider, for example, if the pres­id­ent were to declare martial law over the U.S. border with Mexico, deploy­ing the army to enforce federal immig­ra­tion laws and setting up milit­ary tribunals to prosec­ute alleged viol­at­ors. In that hypo­thet­ical scen­ario, a federal court would likely apply Young­stown to decide whether the pres­id­ent had exceeded exec­ut­ive author­ity.

Under Young­stown, the courts show vary­ing degrees of defer­ence to pres­id­en­tial action, depend­ing on whether the pres­id­ent is acting in accord­ance with or contrary to the will of Congress. Justice Jack­son iden­ti­fied three “zones” into which pres­id­en­tial actions might fall. In the first zone, Congress has author­ized the pres­id­ent’s conduct, entitling it to maximum defer­ence from the courts. The court will uphold the action unless the federal govern­ment, as a whole, lacks the power to act. In the second zone, Congress has said noth­ing on the matter, so courts must search the Consti­tu­tion to find some inde­pend­ent author­ity for the pres­id­ent’s action. In the third zone, the pres­id­ent’s conduct is contrary to laws Congress has passed. These actions are imper­miss­ible unless Congress has over­stepped its own powers.foot­note8_mfkt0d244

Although Young­stown did not address a declar­a­tion of martial law, Justice Jack­son’s concur­ring opin­ion briefly mentioned the concept. Having stated that the Consti­tu­tion’s drafters “made no express provi­sion for exer­cise of extraordin­ary author­ity because of a crisis,” he included the follow­ing caveat in a foot­note: “I exclude, as in a very limited category by itself, the estab­lish­ment of martial law.”foot­note9_1rwui4145  This language acknow­ledges the possib­il­ity that martial law might exist as an emer­gency power, despite the lack of any express provi­sion for it in the Consti­tu­tion. It does not, however, suggest where that power lies, and certainly does not indic­ate that it belongs solely to the exec­ut­ive branch. Nor does it render the three-zone test inap­plic­able in the context of martial law.

How would a unilat­eral pres­id­en­tial declar­a­tion of martial law on the south­ern border fare under Young­stown’s three-zone test? We start with what Congress has said: Congress has legis­lated so extens­ively with respect to the domestic use of the milit­ary — through, for example, the Posse Comit­atus Act, the Insur­rec­tion Act, the Stafford Act, the Non Deten­tion Act, and vari­ous other provi­sions within Title 10 of the U.S. Code — that it has “occu­pied the field.”foot­note10_q35jnyl46  This means that Congress has created such a dense and compre­hens­ive network of rules that anything the pres­id­ent does in this area that is not affirm­at­ively author­ized by stat­ute is almost neces­sar­ily against Congress’s will. Such actions, includ­ing a hypo­thet­ical pres­id­en­tial declar­a­tion of martial law that Congress has not author­ized, would fall within Young­stown’s third zone.

Further­more, the Posse Comit­atus Act creates a general rule that it is unlaw­ful for federal milit­ary forces to engage in civil­ian law enforce­ment activ­it­ies — even if they are merely supple­ment­ing rather than supplant­ing civil­ian author­it­ies — except when doing so is expressly author­ized by Congress.foot­note11_qrfw5er47  As it is gener­ally under­stood, martial law neces­sar­ily involves milit­ary parti­cip­a­tion in civil­ian law enforce­ment. While there are a number of stat­utory excep­tions to the Posse Comit­atus Act, none of them author­izes the pres­id­ent to declare martial law, as part III of this report explains. There­fore, the pres­id­ent’s declar­a­tion of martial law would directly viol­ate the act, which again places it within zone 3 under Young­stown.

In zone 3, the pres­id­ent’s powers are at their “lowest ebb” and pres­id­en­tial actions are rarely upheld.foot­note12_hp27qkg48  This reflects a general rule of consti­tu­tional law: that laws passed by Congress within the scope of its own consti­tu­tional powers “disable” contrary exec­ut­ive action.foot­note13_cxlgku549  In other words, when Congress and the pres­id­ent disagree, Congress wins. Under Young­stown, this rule may be over­come — and the pres­id­ent can act against the will of Congress — only when a “conclus­ive and preclus­ive” grant of pres­id­en­tial power in the Consti­tu­tion author­izes the chal­lenged action.foot­note14_xeijkz950  If the Consti­tu­tion gives Congress any powers in that area, then the pres­id­ent’s author­ity is not “conclus­ive and preclus­ive,” and Congress’s will must prevail.

The crit­ical ques­tion, then, is how the Consti­tu­tion alloc­ates the powers related to domestic deploy­ment of the milit­ary. Three provi­sions of the Consti­tu­tion implic­ate that sort of author­ity: the Call­ing Forth Clause in Article I, the Guar­an­tee Clause in Article IV, and the Commander in Chief Clause in Article II.foot­note15_9i7to1751  The balance of power estab­lished by these provi­sions decis­ively favors Congress over the pres­id­ent.

The Call­ing Forth Clause empowers Congress to “provide for” — that is, to regu­late and control — the author­ity and proced­ures for “call­ing forth the Mili­tia to execute the Laws of the Union, suppress Insur­rec­tions and repel Inva­sions.” As Justice Jack­son explained in Young­stown, this provi­sion was “writ­ten at a time when the mili­tia rather than a stand­ing army was contem­plated as the milit­ary weapon of the Repub­lic.”foot­note16_16g5lbm52  It thus “under­scores the Consti­tu­tion’s policy that Congress, not the Exec­ut­ive, should control util­iz­a­tion of the war power as an instru­ment of domestic policy.”foot­note17_gwr4nnr53

The Guar­an­tee Clause requires the United States to “protect each [state] against Inva­sion; and on Applic­a­tion of the Legis­lature, or the Exec­ut­ive (when the Legis­lature cannot be convened) against domestic Viol­ence.” This language is less clear-cut than the Call­ing Forth Clause, but it certainly does not consti­tute a “conclus­ive and preclus­ive” commit­ment of power to the exec­ut­ive. Instead, it grants author­ity to the federal govern­ment as a whole. Further­more, it only allows unilat­eral federal action in the case of inva­sion. In the event of “domestic viol­ence,” the affected state must request help before the federal govern­ment can act.

Lastly, the Commander in Chief Clause would not enable the pres­id­ent to unilat­er­ally declare martial law in disreg­ard of the Posse Comit­atus Act and other stat­utes that regu­late the domestic use of the milit­ary. To start, the Commander in Chief Clause is not a source of domestic regu­lat­ory author­ity for the pres­id­ent.foot­note18_efu9kxn54  As Justice Jack­son explained in Young­stown, “the Consti­tu­tion did not contem­plate that the title Commander-in-Chief of the Army and Navy” would also mean the pres­id­ent was “Commander-in-Chief of the coun­try, its indus­tries and its inhab­it­ants.”foot­note19_gei6qel55  While the clause grants some­thing more than an “empty title,” its invoc­a­tion does not give the pres­id­ent free rein.foot­note20_8lbn4ao56  In domestic affairs, both gener­ally and with respect to the role of the milit­ary, the Consti­tu­tion envi­sions Congress as the branch in control.foot­note21_rzrpjqz57

To be sure, the pres­id­ent’s powers under the Commander in Chief Clause do not cease to exist inside the territ­orial United States. If a foreign enemy launches a sudden attack inside the United States, it is gener­ally under­stood that the pres­id­ent may act to repel that attack, even if Congress has not given its bless­ing.foot­note22_9rzx8p558  And if Congress has author­ized milit­ary action, the pres­id­ent controls the actual conduct of milit­ary oper­a­tions, even if that fight­ing is taking place within the coun­try’s borders (for instance, during a foreign inva­sion or civil war). But the former power is quite limited, and the latter relies on prior congres­sional author­iz­a­tion.

Because the Consti­tu­tion does not give the pres­id­ent “conclus­ive and preclus­ive” author­ity over the domestic use of the milit­ary — and, on the contrary, expli­citly vests power in the legis­lat­ive branch — the pres­id­ent cannot act against Congress’s wishes in this area. Accord­ingly, a unilat­eral declar­a­tion of martial law by the pres­id­ent today — on the south­ern border or else­where — would not survive a legal chal­lenge under Young­stown.

It bears emphas­iz­ing that this conclu­sion is compelled partly by the Consti­tu­tion and partly by federal law. It is possible that, in the absence of the Posse Comit­atus Act and other laws regu­lat­ing domestic milit­ary activ­ity, the pres­id­ent could rely on some inde­pend­ent exec­ut­ive power to declare martial law. But that scen­ario is hypo­thet­ical and the likely legal outcome is uncer­tain. The real­ity is that the domestic role of the U.S. milit­ary is subject to pervas­ive stat­utory regu­la­tion. This has altered what might other­wise be the scope of the pres­id­ent’s consti­tu­tional author­ity and precludes a unilat­eral pres­id­en­tial declar­a­tion of martial law.foot­note23_wzm23cg59

In the imagined scen­ario described earlier, the pres­id­ent set up milit­ary tribunals to try viol­at­ors of federal immig­ra­tion law. The Posse Comit­atus Act, however, only applies to milit­ary parti­cip­a­tion in law enforce­ment. When it comes to milit­ary involve­ment in judi­cial func­tions, the analysis changes, and the law is char­ac­ter­ized by profound uncer­tainty. While the Call­ing Forth Clause expressly contem­plates the use of milit­ary forces to execute the law, no provi­sion of the Consti­tu­tion author­izes the milit­ary to perform the func­tions assigned to the judi­cial branch under Article III. Nonethe­less, the Supreme Court’s 1866 decision in Ex parte Milligan suggests that the pres­id­ent can, in certain circum­stances, impose martial law in an area and replace the civil­ian courts there with milit­ary tribunals.foot­note24_78z39wt60

The Court’s reas­on­ing in Milligan has some strik­ing incon­sist­en­cies and must be inter­preted cautiously. In one part of the opin­ion, the Court firmly asserts that emer­gency condi­tions can never justify exceed­ing the bounds of the Consti­tu­tion.foot­note25_85o7z2w61  Else­where, however, the Court says that “neces­sity” might warrant declar­ing martial law and using milit­ary tribunals to try civil­ians if regu­lar courts are unavail­able.foot­note26_nt6jjbk62  Import­antly, this latter part of the Court’s opin­ion is dicta, rather than a neces­sary and bind­ing part of the Court’s hold­ing. That alone gives reason to doubt whether there is indeed a “neces­sity” loop­hole that allows the milit­ary to put civil­ians on trial. But the larger issue is that a neces­sity excep­tion to the Consti­tu­tion is impossible: it is a funda­mental prin­ciple of U.S. consti­tu­tional law, reaf­firmed count­less times both before and after the Milligan ruling, that the govern­ment is always constrained by the Consti­tu­tion, no matter the circum­stances.foot­note27_236o5yp63

In any event, even if the neces­sity-based excep­tion artic­u­lated in Milligan were considered to be author­it­at­ive, it would be extremely narrow. It would allow the milit­ary to supplant civil­ian courts only during an actual war, on “the theatre of active milit­ary oper­a­tions,” where the chaos and fight­ing are so severe that the regu­lar courts have been forced to close and are unable to oper­ate.foot­note28_saoc­sjm64

The possib­il­ity of using martial law to replace civil­ian courts with milit­ary tribunals should not be confused with the rule estab­lished by Ex parte Quirin in 1942.foot­note29_qld3u2b65  Quirin and a hand­ful of more recent Supreme Court decisions related to the U.S. milit­ary’s deten­tion facil­ity at Guantanamo Bay allow U.S. citizens not serving in the milit­ary to be tried by milit­ary commis­sion — a partic­u­lar type of tribunal used by the U.S. milit­ary — if they are “enemy combatants.”foot­note30_bexl0gg66  These indi­vidu­als, the Court has held, are subject to the inter­na­tional law of war. As a result, Congress may author­ize their trial by milit­ary commis­sion even if civil­ian courts are open and func­tion­ing, pursu­ant to its author­ity to “define and punish . . . Offences against the Law of Nations.”foot­note31_ucht76067  These decisions are not about martial law. They demarc­ate the line between milit­ary and civil­ian juris­dic­tion, rather than allow­ing the milit­ary to exer­cise juris­dic­tion in an area ordin­ar­ily reserved for civil­ian courts.

No exist­ing federal stat­ute expli­citly author­izes the pres­id­ent to declare martial law.foot­note1_zmzws1e68  However, there are a number of stat­utory excep­tions to the Posse Comit­atus Act that allow the pres­id­ent to deploy the milit­ary domest­ic­ally.foot­note2_ek535za69  The most import­ant of these is the Insur­rec­tion Act.foot­note3_rx8n­m8g70  Rather than a single pack­age of legis­la­tion, the Insur­rec­tion Act consists of a series of stat­utes that were enacted between 1792 and 1871, with a few amend­ments in the 20th century.foot­note4_punfqrk71

Three of the Insur­rec­tion Act’s provi­sions grant the pres­id­ent power to deploy troops domest­ic­ally. The first two, Sections 251 and 252, are relat­ively straight­for­ward and mirror the language of the Call­ing Forth Clause. Under Section 251, if there is an insur­rec­tion in a state, and the state’s legis­lature (or governor, if the legis­lature is unavail­able) requests federal aid, then the pres­id­ent may deploy the National Guard or the regu­lar armed forces to suppress the insur­rec­tion.foot­note5_zb6ib7z72  Section 252 allows the pres­id­ent to deploy troops without a request from the affected state — indeed, even against the state’s wishes — in order to “enforce the laws” of the United States or to “suppress rebel­lion” whenever “unlaw­ful obstruc­tions, combin­a­tions, or assemblages, or rebel­lion” make it “imprac­tic­able” to enforce federal law in that state by the “ordin­ary course of judi­cial proceed­ings.”foot­note6_sy3kk0x73

Noth­ing in the plain language of Sections 251 and 252 indic­ates that they author­ize martial law. The clause in Section 251 that empowers the milit­ary to “suppress an insur­rec­tion” does not suggest that federal troops may take over the role of the civil­ian govern­ment in the process. Rather, it contem­plates that the milit­ary may assist over­whelmed civil­ian author­it­ies by doing exactly what soldiers are trained to do: fight and defeat an armed and hostile group.

Section 252 suggests a more expans­ive power: it allows the milit­ary to enforce federal law, not just to suppress an insur­rec­tion. Nonethe­less, it still does not imply that the milit­ary may push aside the civil­ian author­it­ies. In its 1946 decision in Duncan, the Supreme Court made clear that when a stat­ute author­izes the milit­ary to encroach on the affairs of civil­ian govern­ment, the Court will inter­pret it extremely narrowly. If the stat­ute does not specific­ally say that Congress meant to disrupt the “tradi­tional bound­ar­ies” between civil­ian and milit­ary power, the Court will not imply that intent on Congress’s behalf.foot­note7_cyen­qon74  Because Section 252 does not expressly author­ize the displace­ment of civil­ian author­it­ies, it should not be read as license to turn the normal rela­tion­ship between civil­ian and milit­ary power on its head. Instead, it should be under­stood merely as author­iz­a­tion for the milit­ary to assist civil­ian govern­ment offi­cials when they are over­whelmed by forces attempt­ing to obstruct law enforce­ment and judi­cial proceed­ings.

Section 253 is the only substant­ive provi­sion of the Insur­rec­tion Act that might, on its face, be read to author­ize a limited form of martial law. Among other things, it allows the pres­id­ent to use the National Guard or the active duty armed forces to suppress “any insur­rec­tion, domestic viol­ence, unlaw­ful combin­a­tion, or conspir­acy” in a state if it “so hinders the execu­tion of [state or federal] laws” that “any part or class of [the state’s] people” is deprived of a consti­tu­tional right and “the consti­tuted author­it­ies” of the state “are unable, fail, or refuse to protect” that right.foot­note8_8q5qpm875  As in Section 252, the desires of the state are irrel­ev­ant.

By its express terms, Section 253 contem­plates a situ­ation in which, rather than need­ing help from the milit­ary to enforce the laws, the civil­ian author­it­ies are just not enfor­cing them. As a result, the pres­id­ent may send in troops to suppress whatever insur­rec­tion or other viol­ence is caus­ing a portion of the people in that state to be deprived of a consti­tu­tional right. Thus, if troops are deployed under Section 253, they will, to some extent at least, take over the role of the civil­ian govern­ment. However, the legis­lat­ive history of Section 253 indic­ates that it is best under­stood as allow­ing the milit­ary to supplant only the local police, and to do so in service of laws duly enacted by civil­ian author­it­ies.foot­note9_qg79uqx76  This is so narrow that call­ing it even a “limited form” of martial law would prob­ably be an exag­ger­a­tion. Indeed, the Depart­ment of Defense’s own regu­la­tions emphas­ize the import­ance of main­tain­ing the “primacy of civil­ian author­ity” when troops are deployed in support of law enforce­ment oper­a­tions.foot­note10_7uk0uj477

The Insur­rec­tion Act is an excep­tion to the Posse Comit­atus Act, but there are also some circum­stances in which the latter law simply does not apply. As part V of this report explains, the Posse Comit­atus Act does not affect the abil­ity of states to call up their National Guard forces and deploy them within their own borders. In that situ­ation, National Guard troops are oper­at­ing in State Active Duty status. The Posse Comit­atus Act also does not apply when National Guard forces are activ­ated in what is known as Title 32 status, in which the troops remain subject to state command and control, but are used for federal missions and are typic­ally paid for by the federal govern­ment.foot­note11_08qtzo578

On its face, Title 32 appears to author­ize a fairly limited set of activ­it­ies relat­ing to drug inter­dic­tion and counter-drug activ­it­ies, home­land secur­ity protec­tion, and train­ing exer­cises. In early June 2020, however, U.S. Attor­ney General William Barr put forward a shock­ingly broad inter­pret­a­tion of the section of Title 32 that addresses “required train­ing and field exer­cises” for National Guard forces. Under this section, train­ing may include “oper­a­tions or missions under­taken . . . at the request of the Pres­id­ent or Secret­ary of Defense.”foot­note12_llicxeh79  In a formal letter explain­ing the deploy­ment of several states’ National Guard forces to Wash­ing­ton, DC, during protests that followed the police killing of George Floyd in May 2020, Barr asser­ted that this language allows the pres­id­ent to use the National Guard at any time and for any reason — a read­ing that effect­ively creates a gaping loop­hole in the Posse Comit­atus Act.foot­note13_lwb5arl80

The attor­ney gener­al’s inter­pret­a­tion is suspect for a number of reas­ons, but even if it were correct, it would not author­ize martial law.foot­note14_z241e1o81  As with the Insur­rec­tion Act, there is no clear state­ment in Title 32 to suggest that Congress inten­ded to reverse the usual consti­tu­tional order in which the milit­ary remains subor­din­ate to civil­ian author­ity. Under the Supreme Court’s reas­on­ing in Duncan, the language cited by Barr could at most be read to author­ize the use of the milit­ary to assist civil­ian law enforce­ment author­it­ies.foot­note15_46do2li82

In short, no exist­ing stat­ute author­izes the pres­id­ent to declare martial law. Congress has given the pres­id­ent consid­er­able author­ity, however, to use troops domest­ic­ally to assist in civil­ian law enforce­ment activ­it­ies. Deploy­ing troops under the Insur­rec­tion Act might not raise all of the same concerns that would be asso­ci­ated with a declar­a­tion of martial law, but there is reason to worry whenever a pres­id­ent uses the milit­ary as a domestic police force — partic­u­larly without the consent of the state to which the armed forces are sent.

To start, using the milit­ary to enforce the law flies in the face of Amer­ican tradi­tion. The Founders were deeply suspi­cious of the very idea of a national stand­ing army, believ­ing that it could be used as an instru­ment of oppres­sion and could pose a threat to the autonomy of the indi­vidual states.foot­note16_8kn4g0d83  Many of the protec­tions enshrined in the Consti­tu­tion’s Bill of Rights reflect the hard lessons the Founders learned at the hands of Brit­ish soldiers about the dangers of allow­ing the milit­ary to act as a domestic police force. They worried that a pres­id­ent equipped with a ready and loyal army would be able to subvert demo­cracy and impose his will on the states and the people.

Moreover, even with the best of inten­tions, asking the milit­ary to stand in for the police can yield disastrous results. Soldiers are trained to fight and destroy an enemy, one that gener­ally lacks consti­tu­tional rights. As such, they are poorly suited to perform­ing the duties of police. Forcing them into that role can increase the risk of viol­ence.foot­note17_sogwg4t84  In the words of a Minnesota National Guard member facing deploy­ment in response to the protests that followed the murder of George Floyd: “We’re a combat unit not trained for riot control or safely hand­ling civil­ians in this context. Soldiers up and down the ranks are scared about hurt­ing someone.”foot­note18_sn78×7l85

Even if Congress were to author­ize martial law, and the Supreme Court were to uphold its power to do so, the Consti­tu­tion would still apply. Congress, the pres­id­ent, and the Supreme Court are bound at all times by the Consti­tu­tion and possess only the powers it confers. None of those powers allows the govern­ment to suspend or viol­ate consti­tu­tional rights by martial law or by any other means. On the contrary, as the Supreme Court explained in Milligan, “the Consti­tu­tion of the United States is a law for rulers and people, equally in war and in peace, and covers with the shield of its protec­tion all classes of men, at all times, and under all circum­stances.”foot­note1_ehx4o7x86

The Consti­tu­tion allows Congress to suspend habeas corpus, but every other right it guar­an­tees is inten­tion­ally left “forever invi­ol­able.”foot­note2_mad9bdj87  There is no consti­tu­tional proced­ure for suspend­ing the First Amend­ment’s protec­tion of free expres­sion, the Fourth Amend­ment’s prohib­i­tion on “unreas­on­able searches and seizures,” the Fifth and Sixth Amend­ment rights to trial by jury and the assist­ance of a lawyer, or the Fifth Amend­ment right not to “be deprived of life, liberty, or prop­erty, without due process of law.”

For the most part, however, the rights conferred by the Consti­tu­tion are not abso­lute. When courts decide whether those rights have been viol­ated, they consider the circum­stances —includ­ing not only the impact on the indi­vidual whose rights are affected, but also the interests of the govern­ment — and there is often a balan­cing involved.foot­note3_syx49xx88  In the kind of emer­gency that would justify the declar­a­tion of martial law, the govern­ment might not have to provide the same compre­hens­ive proced­ures required in ordin­ary times before detain­ing someone or confis­cat­ing prop­erty. But the right to due process remains, along with all other consti­tu­tional rights, and the federal courts have the power to decide whether they have been viol­ated.foot­note4_t5qjyw489

The Supreme Court, without actu­ally endors­ing the federal govern­ment’s power to declare martial law, has estab­lished that such declar­a­tions are subject to judi­cial review. At a minimum, in a state or territ­ory that the federal govern­ment has placed under martial law, indi­vidu­als who have been detained by the milit­ary may ask a federal court to order their release by peti­tion­ing for the writ of habeas corpus.foot­note5_x0l884h90  A court that considers an indi­vidu­al’s peti­tion can decide whether the declar­a­tion of martial law was consti­tu­tion­ally permiss­ible.foot­note6_83rcq8791  The review­ing court can also decide whether the milit­ary’s partic­u­lar actions — such as the decision to arrest and detain the person peti­tion­ing for habeas corpus — have viol­ated the Consti­tu­tion or exceeded the powers gran­ted by the stat­ute (if any) that author­ized martial law.foot­note7_3e9f54r92

When determ­in­ing whether the milit­ary has exceeded its stat­utory author­ity, courts will construe that author­ity narrowly. This rule derives from the Supreme Court’s decision in Duncan. Usually, the term martial law refers to the milit­ary taking the place of civil­ian author­it­ies. But in Duncan, the Court held that simply putting the words “martial law” in a stat­ute is not suffi­cient to author­ize such extreme meas­ures. The Court explained that martial law has never been precisely or conclus­ively defined. It emphas­ized the import­ance of the “tradi­tional bound­ar­ies” between milit­ary and civil­ian power in the United States, and iden­ti­fied three basic prin­ciples that make up those bound­ar­ies: first, that the milit­ary must remain subor­din­ate to civil­ian control and the law; second, that it may only assist civil­ian author­it­ies when it is deployed domest­ic­ally; and, third, that it may neither inter­fere with nor usurp the roles of the legis­lature or the judi­ciary.foot­note8_fomk­tr293

The Court concluded that, because these prin­ciples are deeply engrained in U.S. law and insti­tu­tions, any law that purports to abridge them must be clear and specific about how they may be abridged. For instance, a stat­ute might expressly author­ize the use of milit­ary tribunals to try civil­ians in areas where an inva­sion has resul­ted in the clos­ure of the courts and the govern­ment has declared martial law. If that specificity is lack­ing — if the stat­ute or its legis­lat­ive history do not define what Congress meant by martial law — then courts will assume that Congress inten­ded only to author­ize the milit­ary to “act vigor­ously for the main­ten­ance of an orderly civil govern­ment and for the defense of the [area] against actual or threatened rebel­lion or inva­sion.”foot­note9_zku45xu94

The law govern­ing states’ use of martial law is compar­at­ively simple. The Supreme Court has expressly held that indi­vidual states have the power to declare martial law.foot­note1_phwonns95  The Posse Comit­atus Act does not apply to the states and thus does not affect their abil­ity to use their milit­ary forces for domestic law enforce­ment.foot­note2_lxufms296  Instead, a state declar­a­tion of martial law is valid simply if it is author­ized by the consti­tu­tion or laws of the state.foot­note3_ckmsi7997

But this does not mean that states can do whatever they want. The Suprem­acy Clause estab­lishes that the U.S. Consti­tu­tion, federal laws, and treat­ies consti­tute “the supreme Law of the Land,” mean­ing that states have no author­ity to over­ride them.foot­note4_qo1wbta98  The Supreme Court has emphas­ized that even under martial law, state offi­cials are bound both by the U.S. Consti­tu­tion and by valid federal laws.foot­note5_1n4gyk499  If indi­vidu­als wish to chal­lenge a state declar­a­tion of martial law, they may seek injunct­ive relief in federal court — or, if they have been detained, they may peti­tion for the writ of habeas corpus.foot­note6_3py0f60100  The review­ing court will show great defer­ence to a state governor’s determ­in­a­tion that the situ­ation required a declar­a­tion of martial law.foot­note7_qkzt4nz101  However, the court will exam­ine whether any milit­ary actions taken under that declar­a­tion have run afoul of the Consti­tu­tion or federal law.foot­note8_ypdkooy102

On that ques­tion, the Supreme Court ruled in Moyer v. Peabody and Ster­ling v. Constantin that actions taken under state martial law are valid if they are “conceived in good faith, in the face of the emer­gency, and [are] directly related to the quelling of the disorder or the preven­tion of its continu­ance.”foot­note9_rgu96qw103  This “good faith/direct rela­tion­ship” stand­ard could be read to imply that state offi­cials may take actions under martial law that would in other circum­stances viol­ate the Consti­tu­tion. In other words, it suggests that in an emer­gency, good faith and a direct rela­tion­ship between means and ends are enough to satisfy the Due Process Clause of the Four­teenth Amend­ment.

However, the “good faith/direct rela­tion­ship” stand­ard from Ster­ling and Moyer is more appro­pri­ately read as a floor than as a ceil­ing. Both cases certainly suggest that a state’s actions would be invalid if they were made in bad faith or were not directly related to quelling the disturb­ance that precip­it­ated them. But it is not clear that good faith and a direct rela­tion­ship alone would be enough to satisfy due process or other consti­tu­tional require­ments today. Moyer and Ster­ling are very old cases. Decided in 1909 and 1932, respect­ively, they pred­ate virtu­ally all of the Supreme Court’s modern due process and civil rights case law.foot­note10_cdfi0u5104  Under those preced­ents, it is likely that the Court would apply a more demand­ing stand­ard of review.

The law govern­ing martial law is complic­ated and unsettled, but a few prin­ciples can be found in the Supreme Court’s rulings on this topic and on the limits of exec­ut­ive power more gener­ally. The Consti­tu­tion gives Congress author­ity to regu­late the domestic deploy­ment of the milit­ary, and Congress has enacted compre­hens­ive legis­la­tion in that area. Because that legis­la­tion does not include author­iz­a­tion for the pres­id­ent to impose martial law, the pres­id­ent has no power to do so. Even if Congress were to provide author­iz­a­tion, the Supreme Court has not conclus­ively decided that the federal govern­ment is consti­tu­tion­ally empowered to declare martial law. But it is clear that such a power, if it exis­ted, would have to be exer­cised within the bounds of the Consti­tu­tion, and the milit­ary’s actions would be subject to judi­cial review.

The state martial law power is more clearly estab­lished, but there are signi­fic­ant limits. States may declare martial law whenever it is author­ized by state law, and federal courts are likely to defer to a state governor’s decision that doing so was neces­sary. However, the Consti­tu­tion and valid federal laws will still constrain states’ conduct under the declar­a­tion, and judi­cial review will be avail­able in federal court.

The exact scope and limits of martial law will thus remain danger­ously unclear until Congress and state legis­latures enact new laws that better define them.

Beyond this, the Supreme Court preced­ent is too old, sparse, and incon­sist­ent to provide any certainty around martial law. Indeed, in the absence of legis­la­tion specific­ally address­ing martial law, even the above prin­ciples are subject to compet­ing inter­pret­a­tions and would likely be disputed by exec­ut­ive offi­cials seek­ing to use milit­ary forces more aggress­ively. The exact scope and limits of martial law will thus remain danger­ously unclear until Congress and state legis­latures enact new laws that better define them.

Moreover, even without the power to declare martial law, the pres­id­ent still has extens­ive author­ity to deploy the milit­ary domest­ic­ally to perform law enforce­ment func­tions. The Insur­rec­tion Act — and possibly Title 32 as well — leave it almost entirely up to the pres­id­ent to decide when and where to use U.S. armed forces at home against Amer­ic­ans. To some observ­ers, a deploy­ment of troops under the Insur­rec­tion Act might look and feel very much like martial law. Given the degree of confu­sion over the term, some within the media or the govern­ment itself might even call it martial law. Although that label would be inac­cur­ate and the milit­ary’s author­ity would be substan­tially less extens­ive than under martial law, the fact remains that any use of U.S. armed forces as a domestic police force repres­ents a depar­ture from Amer­ican tradi­tion and carries inher­ent risks. Thus, as with martial law, Congress urgently needs to pass legis­la­tion clari­fy­ing and limit­ing the pres­id­ent’s powers under the Insur­rec­tion Act and Title 32.

The author would like to thank the Bren­nan Center’s Eliza­beth Goitein for her invalu­able guid­ance, comments, and sugges­tions, as well as Mireya Navarro, Tim Lau, Stephanie Sykes, Jeanne Park, Matthew Harwood, Ryan Witcombe, and Justin Charles for their outstand­ing commu­nic­a­tions support. Sahil Singhvi, Victoria Ochoa, Rachel Lesser, Aleena Nasir, Amelia Shapiro, and Sheel Patel deserve special thanks for their dili­gent research efforts, which made this report possible.

The author would also like to thank profess­ors Steve Vladeck, Bill Banks, and Jeff Powell. The know­ledge and insight they contrib­uted through conver­sa­tions and corres­pond­ence greatly benefited the report.

The Bren­nan Center grate­fully acknow­ledges The Bauman Found­a­tion, CS Fund/Warsh-Mott Legacy, Demo­cracy Fund, The Endeavor Found­a­tion, Inc., The William and Flora Hewlett Found­a­tion, and Open Soci­ety Found­a­tions for their gener­ous support of our work.