Jurisdictional forum shopping

Why do companies incorporate in a state where they keep no offices and no staff?

There is a single low building in Wilmington that serves as the registered address of hundreds of thousands of companies. None of them operates there. Delaware has fewer than a million residents, yet a large majority of Fortune 500 companies — the figure usually quoted is around two thirds — are incorporated under its law, and the number of registered entities in the state exceeds the number of people.

The obvious guess is tax avoidance. But a company headquartered in California still pays California tax on what it earns in California, whatever its charter says. So if the money does not move, what does?

Start by separating two things that ordinary speech runs together. Where a company is — its offices, staff, customers, taxable activity — is a fact about the physical world. Where a company is chartered is a fact about which sovereign's law defines it. Are those necessarily the same place?

American law says no, and says so explicitly. Under the internal affairs doctrine, the law governing a corporation's internal affairs — what directors owe shareholders, how a merger is approved, when a board may resist a takeover — is the law of the state of incorporation, regardless of where the corporation actually does business. That single rule is the whole precondition for what follows. It decouples the legal home from the physical one, and it means the choice of legal home can be made once, on paper, and carries almost no operational cost.

Now we have the shape of a gradient problem. Whenever a quantity can move freely and the destinations differ in some value, flow follows the difference. So: what differs between states, and in which direction does it point?

Not the statutes alone — other states have copied Delaware's corporate code almost verbatim, and that copying did not move the companies. Ask what could not be copied. Delaware's Court of Chancery is a specialist equity court with no juries, judges chosen for corporate expertise, and a practice of deciding fast enough to matter in a live deal. Behind it sits well over a century of accumulated decisions on almost every question a board might face. A general counsel choosing a charter is buying predictability: a fairly reliable answer, in advance, about how a dispute would go. That stock of precedent is the part a rival legislature cannot enact into existence.

Here is the step that makes the pattern stable rather than temporary. In an ordinary gradient, flow erodes the difference until it flattens. Why doesn't this one flatten? Because the destination is paid to maintain it. Delaware collects franchise taxes and entity fees that supply on the order of a third of its general revenue, and it reinvests in the thing that attracts the flow — keeping the code current, the court staffed, the registry fast. Each new incorporation also adds to the case law, which increases predictability, which attracts the next one. The gradient is fed by its own flow.

Whether that competition is good is genuinely contested and has been since the 1970s: one side, following William Cary, reads it as a race to the bottom in which states sell management protection against shareholders; the other, following Ralph Winter, argues that capital markets punish charters that hurt shareholders, so the race runs upward. The evidence has never fully settled it.

Three refinements matter.

First, this is not evasion of substantive obligation. A Delaware charter does not exempt a company from the employment law, environmental law, or taxes of the states where it operates. What it selects is the body of law governing the relationship between the company and its own shareholders and directors — a narrower thing than "escaping regulation," and the narrowness is why the practice is legal and uncontroversial as a matter of doctrine.

Second, the same gradient logic runs in litigation, not just chartering, and the flow there is more contested. Plaintiffs choose among venues where jurisdiction is proper, and concentrations follow — patent suits famously piled into one Texas district until the Supreme Court's 2017 TC Heartland decision tightened venue and redirected much of the flow toward Delaware; large bankruptcies cluster in a handful of courts. The mechanism is identical: an open path plus a perceived difference in outcomes.

Third, the flow responds to the gradient as perceived, not as it truly is. Much of Delaware's advantage now is that everyone — underwriters, investors, law firms — already knows Delaware law, so choosing it needs no explanation to anyone. That is a coordination effect layered on top of the legal one, and it would keep the flow running for a long while even if the underlying legal advantage narrowed.

• Whether a federal corporate charter would remove the gradient or simply relocate it.
• How the European Union's freedom of establishment cases produced a comparable, if weaker, flow toward permissive company-law regimes.
• The empirical work testing whether Delaware incorporation is associated with higher firm value, and why it is so hard to separate cause from selection.

Every term the collection defines is gathered in the glossary.