Concept
Citizenship of Corporations
Diversity jurisdiction
Citizenship of Corporations
A corporation has no domicile in the sense of a home it intends to return to, so the presence-plus-intent test that sorts individuals cannot be applied to it. Congress filled the gap with a statutory definition. Under 28 U.S.C. § 1332(c)(1), a corporation is deemed a citizen of every State and foreign state by which it has been incorporated and of the State or foreign state where it has its principal place of business.1 The lead-in to subsection (c) makes the definition govern both § 1332 and the removal statute, § 1441, so the same count applies whether a plaintiff files in federal court or a defendant removes from state court.2 This section takes the rule in four steps: the dual citizenship the statute creates, the nerve-center test that fixes the principal place of business, the special rule for liability insurers, and the effect of foreign incorporation or headquarters.
Dual Corporate Citizenship
The statute gives a corporation two kinds of citizenship at once, and the opposing party must be diverse from all of them. A corporation is a citizen of every State in which it is incorporated, and of the one State where it has its principal place of business. The pleader has no option to choose whichever of those States helps; the corporation is regarded as a citizen of both, so diversity is lacking if any adverse party is a citizen of either.3 A corporation incorporated in Delaware with its principal place of business in Ohio therefore cannot be sued in federal court on diversity grounds by a Delaware citizen or by an Ohio citizen.
The reason for counting two States lies in how the law developed. For roughly a century the federal courts treated a corporation as a citizen only of the State that chartered it, and by 1928 that rule was close to absolute, so a company could rest on a charter from a State where it did no business at all.4 A company could then pick a charter State to suit its litigation preferences, which cut against the purpose of diversity jurisdiction, namely shielding out-of-state parties from local prejudice. In 1958 Congress kept the incorporation test and added the principal-place-of-business State alongside it.5 Suppose Silver Mining is chartered in Delaware but all of its mines and workers are in Nevada. Under the old rule it was an outsider to Nevada citizens, even though it was local there in every practical sense. Under the statute it is a citizen of both Delaware and Nevada, so a Nevada plaintiff and a Delaware plaintiff each lack diversity from it.6
Two features of the rule are easy to miss. First, the incorporation half applies even when the corporation has no presence in the charter State, because the act of incorporation is itself the tie the statute counts.7 Second, a corporation has only one principal place of business. The statute refers to the State where it has its principal place of business, and a company that does heavy business in several States is still a citizen only of the one that qualifies. If Silver Mining's principal place of business is Nevada and it also operates a mine in Wyoming, it is diverse from a Wyoming citizen.8
The application pair shows which fact does the work. A Delaware corporation headquartered in New Jersey that does most of its business in California is a citizen of Delaware and New Jersey, so a California plaintiff is diverse from it. Change one fact so that the corporation is headquartered in California, and the same California plaintiff now faces a California citizen and diversity fails. Complaints must be pleaded accordingly: an allegation that a corporate party is a citizen of State X will not do, and the pleading must state where the corporation is incorporated and where it has its principal place of business.9
The Nerve-Center Test
Because the principal place of business decides one of the corporation's citizenships, courts need a rule for locating it. For decades the circuits disagreed. Some looked to the corporation's home office, some to day-to-day control, and some to the State where the bulk of its activity occurred.11 In Hertz Corp. v. Friend, two California citizens sued Hertz in California state court on wage-and-hour claims, and Hertz removed on the ground that it was a citizen of Delaware and New Jersey, where its headquarters sat. The plaintiffs answered that California, with the largest share of Hertz's rental locations, employees, and revenue, was its principal place of business.12 The Supreme Court held that a corporation's principal place of business is the place where its high-level officers direct, control, and coordinate its activities, the "nerve center," which will usually be its headquarters.13
The Court gave three reasons, and they explain why the test is shaped as it is. The first is the statutory text. The word "place" is singular, and "principal" calls for the main or leading place. Because "place" follows "State where," it also means a place within a State rather than the State itself. A headquarters is a single place within a State, whereas a measurement of total business activity by State looks at the State as a whole.14 The second reason is administrative simplicity. A jurisdictional test that requires weighing plants, payroll, and sales eats up time and money that the parties should spend on the merits, and it produces appeals and gamesmanship.15 The third reason comes from the drafting history, in which the Judicial Conference dropped an earlier gross-income proposal because critics found it too complex, which suggests the final words should be read to be no more complicated than that test.16 The Court also saw a practical danger in the activities test: if a corporation's activities simply track a State's population, nearly every national retailer would be a California citizen.17
The test has limits that matter on an exam. The headquarters must be the actual center of direction, control, and coordination, and not merely an office where directors and officers travel for board meetings.18 The party invoking jurisdiction bears the burden of proving the nerve center with competent evidence, and a bare listing of principal executive offices on an SEC filing is not enough. If the record shows that the claimed nerve center is only a mail drop box, a bare office with a computer, or the site of an annual executive retreat, the court looks instead to the place of actual direction and control.19 That rule closes the loophole that incorporation-State shopping once opened, since a company cannot simply declare a convenient headquarters. The Court also conceded that telecommuting and divided command structures will produce hard cases, and that the test can yield results that seem to cut against diversity's purpose. If most visible business activity is in New Jersey but the top officers sit in New York, the principal place of business is New York, and the Court accepted such anomalies as the price of a clear rule.20
Two applications show the test at work. Suppose an airline incorporated in Massachusetts and headquartered there has its reservation center and most employees in Arizona. It is a citizen of Massachusetts only, so an Arizona plaintiff is diverse from it, because the heavy Arizona activity does not make Arizona the principal place of business.21 Now take a bar chain incorporated in Delaware with eight bars in Massachusetts, a few elsewhere, and headquarters in Rhode Island. It is a citizen of Delaware and Rhode Island, and a Massachusetts plaintiff is diverse from it, even though most of its sales occur in Massachusetts.22 The same logic governs removal. A corporate defendant sued in State Green may remove a diversity case only if it is not a citizen of Green, and because Green is not its nerve center or charter State, a large factory there does not block removal.23
Insurers in Direct Actions
Section 1332(c)(1) contains one exception to the ordinary corporate count. In a direct action against the insurer of a liability policy, where the insured is not joined as a defendant, the insurer is also deemed a citizen of the insured's State.25 The insurer keeps its own incorporation and principal-place-of-business citizenships and adds the insured's.26
The reason is a docket problem that Louisiana law created. That State permitted a plaintiff to sue the liability insurer directly without joining the insured, and plaintiffs' lawyers preferred federal court there. A routine accident case between two Louisiana citizens could be turned into a diversity case simply by suing an insurer chartered and based elsewhere.27 Giving the insurer the insured's citizenship puts the real parties on the same footing as if the insured had been sued. Suppose a Louisiana driver injures a Louisiana pedestrian, and the pedestrian sues only the driver's Connecticut-based insurer. The insurer is a citizen of Connecticut and of Louisiana, so the pedestrian is not diverse from it. If the pedestrian sues the driver, or joins the driver as a defendant, the rule is not triggered and the ordinary analysis applies. The Supreme Court has also held that the exception does not apply when the insurer itself is the plaintiff.28
Foreign Corporations
The statute's own text covers foreign corporations, because it counts every State and foreign state of incorporation and the State or foreign state of the principal place of business.30 Before Congress clarified the wording in 2011, courts split over whether § 1332(c) reached companies chartered abroad. Some held that a foreign-chartered corporation was a citizen only of its country of incorporation, which let a company headquartered in New York sue a New York citizen simply because it was incorporated in Liberia rather than Delaware.31 The amended statute removes that anomaly. Every corporation is a citizen of both its place of incorporation and its principal place of business, whether those places are inside or outside the United States.32
Two consequences follow. A foreign corporation with its principal place of business in a particular State cannot invoke diversity against a citizen of that State, and it cannot be sued on diversity grounds by one. And a citizen of a foreign country who sues a United States corporation whose principal place of business is abroad has produced a suit between two aliens, which falls outside diversity; state court remains available in both situations.33
Dual Corporate Citizenship
Section 1332(c)(1) gives a corporation more than one citizenship at the same time, and the party on the other side of the case must be diverse from every one of them. The statute deems a corporation a citizen of "every State and foreign state by which it has been incorporated and of the State or foreign state where it has its principal place of business."35 The list a court builds is therefore every place of incorporation plus one principal-place location. The pleader cannot pick whichever of those States produces diversity, because the corporation is regarded as a citizen of all of them together.36 A corporation chartered in Delaware with its principal place of business in Ohio is thus a citizen of both, and diversity is lacking if any adverse party is a citizen of either.37
The two-part count makes sense once you see what the single-charter rule allowed. For most of the nineteenth century the federal courts treated a corporation as a citizen of the State that chartered it, and by 1928 the Court had made that rule close to absolute. In the Black and White Taxicab decision a company closely identified with one State was allowed into federal court there because it had filed its incorporation papers in another State, perhaps one where it did no business at all.38 Filing a charter is cheap and the choice of State is the company's own, so a business could manufacture or avoid diversity simply by picking where to incorporate. That result undercut the purpose of diversity jurisdiction, which is to protect out-of-state litigants from local bias. The remedy that emerged from the Judicial Conference's study of the problem was a simpler formula than its first proposal, which would have turned on gross income. It kept the State of incorporation and added the State of the principal place of business.39 A company chartered in Delaware but operating entirely in Nevada, such as an invented Silver Mining Company, had been an outsider to Nevada citizens even though it was local there in every practical sense. The statute makes it a citizen of Nevada as well as Delaware and removes that anomaly, while the incorporation tie remains as a second connection.40
The count is asymmetric. A corporation has only one principal place of business, because the statute speaks of "the State or foreign state" where it is located, so heavy activity in other States does not add citizenships.41 Incorporation, by contrast, can occur in several States, and the statute counts each of them. Before the 1958 amendment, a corporation chartered in two States was treated as a citizen only of the State where the suit was brought, so diversity could exist in one federal court and vanish in another on identical facts.42 The current text avoids that arbitrary result by looking at all the States of incorporation at once. A corporation chartered in New York and Massachusetts and headquartered in Texas is therefore a citizen of three States, and a plaintiff from any of them lacks diversity from it.
The same list can include a foreign state. Courts once disagreed over whether § 1332(c) reached a corporation chartered abroad, and some treated such a company as a citizen only of its country of incorporation. That let a company headquartered in New York invoke diversity against a New York citizen because it was incorporated in Liberia rather than Delaware.43 Congress resolved the question in 2011 by providing that every corporation is a citizen of both its place of incorporation and its principal place of business, whether those places are inside or outside the United States.44 Two results follow. A foreign corporation headquartered in a State cannot invoke diversity against that State's citizens, and cannot be sued on that basis by them. A foreign citizen who sues a United States corporation with its principal place of business abroad has sued another alien, which falls outside diversity.45
A matched pair shows which fact decides the outcome. Suppose a corporation incorporated in Delaware has its headquarters in Nevada and also runs a mine in Wyoming. A Wyoming plaintiff is diverse from it, because Wyoming is neither its State of incorporation nor its principal place of business.46 A Delaware plaintiff is not diverse, because of the charter, and a Nevada plaintiff is not diverse, because of the headquarters. Nothing about the corporation's conduct changed between the first case and the other two. Only the plaintiff's home State moved from outside the list to inside it.
The Nerve-Center Test
Section 1332(c)(1) makes a corporation a citizen of the State where it has its principal place of business, but the statute never says how to find that place, and the lower courts answered the question in different ways. The dispute in Hertz Corp. v. Friend shows what hung on the choice. Two California citizens sued Hertz in a California state court on a class claim under California's wage and hour laws, and Hertz tried to remove the case to federal court on diversity grounds.48 The district court applied Ninth Circuit precedent, which counted a corporation's business activity State by State and treated a State with a significantly larger share as the principal place of business. It found that the plurality of Hertz's activities was in California, held Hertz to be a California citizen, and sent the case back to state court.49
The Supreme Court unanimously rejected that approach and adopted a different one. A corporation's principal place of business is the place where its high-level officers direct, control, and coordinate its activities, which the lower courts call the "nerve center," and which will usually be the corporation's headquarters.50 The headquarters counts only if it is the actual center of direction and control. An office used mainly for board meetings, to which directors and officers travel for the occasion, does not qualify.51 On the undisputed facts, Hertz's officers directed the company from New Jersey, so its nerve center was there and not in California, whatever share of its rentals took place in California.52
The Court gave reasons for choosing this test over a count of business activity, and each reason helps you apply the rule to new facts. The first comes from the text. The statute speaks of a single "place" and asks for the principal one, and because "place" follows the words "State where," it must be a location within a State and cannot be the State itself. A headquarters is one identifiable location, whereas a business-activities test measures a whole State's share of the company's operations.53 The second reason is administrative. A court must satisfy itself that it has subject-matter jurisdiction even when no party objects, and a test that requires weighing plants, payroll, and sales makes the parties spend time and money fighting over which court should hear the case rather than the merits. Such a test also generates appeals and invites gamesmanship, because each side can shape the evidence about where activity is concentrated.54 The third reason is the drafting history. The Judicial Conference first proposed a numerical gross-income test and abandoned it after critics found it too complex, which suggests that the final words should be read to be no harder to apply than that rejected test.55
The Court also accepted that the test has costs. Some corporations split their command functions among officers working in several places, and those cases will be hard, although the test at least points courts toward one question, the center of overall direction.56 More troubling is that the result can run against diversity's purpose of shielding out-of-state parties from local prejudice. If the public sees a company's business in New Jersey but its top officers sit across the river in New York, the principal place of business is New York. The company may then remove a New Jersey state case to federal court but cannot remove one filed in New York.57 The same oddity appears when a company headquartered in Illinois is the largest employer in a small Mississippi town. Local residents know it well, yet it counts as a citizen of Illinois, where few people think of it as local, and not of Mississippi.58 The Court said expressly that it must accept such anomalies, because occasional counterintuitive results are the price of a clear and uniform jurisdictional rule.59
The rule also controls who must prove what. The party invoking diversity jurisdiction carries the burden of persuasion, and when the facts are challenged it must support them with competent proof. The Court refused to treat a corporation's listing of its principal executive offices on a securities filing as enough by itself, since that would let a company manufacture jurisdiction by naming a convenient address. If the record shows that the claimed nerve center is a mail drop box, a bare office with a computer, or the site of an annual executive retreat, the court looks to the place of actual direction and control.60
Two invented-looking exam patterns show the steps at work. An airline incorporated and headquartered in Massachusetts that runs its reservation center and employs most of its staff in Arizona is a citizen of Massachusetts alone, so an Arizona plaintiff is diverse from it.61 Likewise, a manufacturer incorporated and headquartered in Red that operates its main plant in Green is not a citizen of Green. It can therefore remove a diversity suit filed against it in Green's state court, even though a Green plaintiff would otherwise raise doubts.62 Change one fact and the answer reverses: if the company's top officers actually ran the business from Green, Green would be its nerve center, and the same plaintiff would destroy diversity and the same removal would fail.
Insurers in Direct Actions
The ordinary corporate count, a citizen of every State of incorporation plus one principal-place State, has a single exception for liability insurers. Some States let an injured person sue the insurer of the wrongdoer directly, instead of suing the wrongdoer and collecting on the judgment from the insurer afterward. Such a suit is called a direct action. When one is brought against the insurer of a liability policy and the insured is not joined as a defendant, § 1332(c)(1) deems the insurer, whether incorporated or unincorporated, a citizen of three sets of places: every State and foreign state of which the insured is a citizen, every State and foreign state in which the insurer is incorporated, and the State or foreign state where the insurer has its principal place of business.64 The insurer therefore keeps its two normal corporate citizenships and takes on the insured's citizenship as well, so the plaintiff must be diverse from all of them.
The reason for the extra citizenship lies in what a direct action really is. The injured plaintiff's actual dispute is with the insured, who caused the harm, and the insurer is in the case only because it must pay if the insured is liable. If the insurer's own citizenship were the only one counted, a plaintiff could turn a quarrel between two local residents into a diversity case just by suing the out-of-state insurer alone. Attributing the insured's citizenship to the insurer closes that route. Congress added the rule in 1964 after Louisiana plaintiffs, who preferred federal court and could sue insurers directly under state law, had been converting routine accidents between two Louisiana citizens into diversity actions and crowding the federal dockets there.65
A matched pair shows which fact controls. Suppose an injured Texan sues a Connecticut insurer directly, without joining the Texan who held the policy. The insurer is a citizen of Connecticut by incorporation and headquarters and also of Texas through the insured, so the Texan plaintiff shares a citizenship with it and diversity fails.66 Now change one fact: the plaintiff joins the Texas insured as a defendant. The exception applies only when the insured is not a party, so the insurer is again treated like any other corporation, a citizen of Connecticut alone. Diversity still fails in that case, but for a different reason, because the Texan insured is now a defendant who shares the plaintiff's State. The statute also reaches only suits against the insurer. The Supreme Court has held that, given its unambiguous language, it does not apply when the insurer is the one bringing the suit.67