Apple spent fifteen years insisting that its tightly closed iPhone, one app store, one payment system, one set of rules enforced by one company, was not a business model but a safety measure, the only way to protect users from the chaos and malware of a more open system. This fortnight the European Union forced the wall open, requiring Apple to permit rival app stores and payment systems, and the fight is being framed, by Apple most of all, as security against freedom, the safe garden against the dangerous wild. That framing is not wrong, exactly, but it misses the deeper question, which an economist named half a century ago and which cuts closer to what is actually at stake: the question of what a user can do when the only door is locked.
Albert Hirschman, in a small and brilliant book, identified the two fundamental responses available to anyone dissatisfied with an organisation, a product, or a state: exit, leaving for an alternative, and voice, staying and complaining, protesting, demanding change. His insight was that these two mechanisms discipline the powerful in different ways, that exit works through the market (lose enough customers and you reform or die) while voice works through politics (hear enough protest and you respond), and that the health of any system depends on at least one of them being available and effective. A closed platform is, precisely, a system engineered to eliminate exit: once you have bought the phone and the apps and the ecosystem, the cost of leaving is enormous, and Apple's genius was to make exit so expensive that it effectively vanished as a discipline, leaving users locked in a garden they could not leave. And when exit is foreclosed, Hirschman's framework says, only voice remains, and voice against a private company is famously weak, because you cannot vote in its board and it need not listen. What the European regulation does, in Hirschman's terms, is restore exit where the platform had abolished it, and thereby restore a discipline on Apple that the closed garden had removed.
The counterargument, and Apple makes it with real force, is that the wall genuinely does protect users, that the App Store's curation really does keep out the malware and the scams that plague more open systems, and that most users neither want nor benefit from exit, they want safety, which the garden provides and the wild does not, so that forcing the garden open serves the sophisticated minority who want alternatives at the cost of the majority who wanted protection. There is truth here. The curation is real, the safety is real, and a certain kind of user is genuinely better off inside the wall. But notice what the argument assumes and Hirschman exposes: that the company which built the wall can be trusted to keep the garden in the users' interest rather than its own, and the whole point of exit is that it is the mechanism which makes that trust unnecessary, which disciplines the gardener regardless of his intentions by giving the users somewhere else to go. A garden with no exit requires you to trust the gardener; a garden with an exit does not, and the difference between those two is the difference between a safety that depends on the goodwill of the powerful and a safety that does not.
Because that is what the fight is actually about, beneath the security framing, and it is why Brussels was right to force the gate. The question is not whether the garden is safe today; it may well be. The question is what disciplines the gardener when his interest and the users' diverge, as it eventually will, and the answer, for fifteen years, was nothing, because exit had been engineered away and voice was toothless against a trillion-dollar company. The regulation does not tear down the wall; it puts a gate in it, and the gate matters not because most users will walk through it but because its mere existence changes the gardener's incentives, restores the discipline of exit, and converts a safety that depended on Apple's goodwill into a safety that Apple must now compete to provide. Apple says the gate makes users less safe. Hirschman's framework suggests the opposite: that a user who can leave is a user who need not trust, and that the freedom to exit is not the enemy of safety but its most reliable guarantor, because it is the one discipline that works whether or not the gardener means well.