Checks on power

This lesson tries to do something uncommon in public discussion: present the actual menu of mechanisms that have historically constrained concentrated power, without dismissing any of them on partisan grounds. Some of the items on this list are politically left-coded; some are right-coded; most are coalitions of conditions rather than single points. The framework is functional, not ideological — what has actually constrained power, in what circumstances, with what costs.

The framing question: if you genuinely worry about the possibility of an over-mighty state, an over-mighty corporation, or an over-mighty individual — and history suggests you should — what tools are on the table?

The empirical question, honestly

Most lists of "checks on power" treat them as obvious goods that should be maintained. The harder analytical question is: which checks actually work, against what kinds of power, with what side effects?

Independent courts. Strong empirical record of constraining state action where the rule of law tradition is established. Weaker where it isn't. Famously vulnerable to slow capture (judicial appointments over decades). Mostly works in democracies; mostly fails in autocracies regardless of formal structure.

Free press. Adversarial press has historically forced major reforms (Pentagon Papers, Watergate, abuses at Abu Ghraib, the 2008 financial reporting that drove the post-crisis reforms). Currently undergoing the most severe structural stress in modern history — collapse of business model, consolidation of ownership, regulatory pressures across democracies. Hard to assess in real time how much it still constrains.

Federalism. Real check in the US specifically — states resisting federal policy is a constant pattern (sanctuary cities, cannabis legalization, climate policy in California, the 2020 election administration). Many federalist systems globally are weaker than the US version. Costs: regulatory inconsistency, race-to-the-bottom dynamics in some areas, mobility-of-capital arbitrage.

Free exit (mobility). Perhaps the most underrated check. When people and capital can leave a regime, the regime has incentive to provide value. The collapse of Soviet bloc was partly a competition the East lost on this dimension. Modern "vote with your feet" patterns (state-to-state migration in the US, country-to-country movement in Europe) constrain governments at the margin. The check fails completely against governments that prevent exit (Soviet Union, modern North Korea, in lesser degree China).

Strong encryption. The most novel check in this list. Effective against bulk surveillance (which is why governments globally have been trying to restrict it). Less effective against targeted surveillance. The Apple-FBI fight, the UK Online Safety Act\'s encryption provisions, and the EU Chat Control proposals are all attempts to weaken this check. The technology side has been winning so far.

The Second Amendment / armed citizenry. Genuinely contested. The framers explicitly intended it as a check against tyrannical government, drawing from English Whig political philosophy and the experience of standing armies. The modern empirical question is whether civilian arms in fact constrain a state that controls drones, intelligence services, and a professional military with armored vehicles. Cases cited in defense of the argument: Ukrainian and Afghan resistance against larger forces; the practical difficulty of imposing martial law in a society with hundreds of millions of firearms. Cases cited against: the asymmetry of modern state power, the costs of an armed society in terms of accidents and interpersonal violence, the fact that the citizen-arms check has not visibly prevented expansions of state power within the US itself. This isn\'t a question with a clean empirical answer; intelligent people disagree.

Civil society / voluntary associations. De Tocqueville\'s great observation about American democracy was the density of voluntary associations — churches, unions, fraternal orders, professional societies, civic groups. These were the substrate of self-governance. The decline of this substrate (Robert Putnam\'s "Bowling Alone") is one of the cleanest documented changes in American civic life over the past 50 years, and one of the harder problems to address.

Markets and competition. Antitrust enforcement, competition law, ease of new business formation. A genuine check on corporate concentration — when applied. The post-1980 weakening of US antitrust enforcement and its recent partial reversal under the current FTC/DOJ leadership is one of the most consequential policy shifts of the past decade. Markets check each other; market concentration removes the check.

Whistleblowers and leakers. Snowden, Manning, the Pentagon Papers (Ellsberg), the Pandora Papers leakers, the FinCEN Files. Without these, the public would know vastly less about what its institutions do. They are imperfect mechanisms — sometimes the leaker has bad motives, sometimes the leak causes harm beyond the disclosed wrongdoing — but they are functionally irreplaceable. The decade-long trend of harsher prosecutions of national security leakers (under both parties) is a structural concern.

Jury nullification. The historic right (and ongoing legal reality) of juries to refuse to convict despite the law. The Fugitive Slave Act prosecutions were widely nullified by Northern juries. Modern juries have nullified marijuana and assisted-suicide cases. It is a thin check, but it is a check.

Elections, parliaments, opposition parties. The classical democratic checks. Working partially, under stress in most developed democracies. Empirical literature suggests competitive elections do constrain incumbents on average; the constraint is weaker where election integrity itself is contested.

The system view

Each check works against some kinds of power and not others. Courts constrain executive action; they don\'t constrain platform algorithms. Encryption constrains bulk surveillance; it doesn\'t constrain financial concentration. Federalism constrains national policy; it doesn\'t constrain transnational corporations. The Second Amendment, whatever you think of it, doesn\'t constrain TSMC.

The Madisonian insight is that you want multiple overlapping checks, not one. A single check that "should be sufficient" is, in practice, a check that some sufficiently determined actor will eventually figure out how to disable. Five overlapping checks, each individually weak but mutually reinforcing, are harder to defeat. The redundancy is the point.

The contemporary concern is that multiple checks are weakening simultaneously. The press is consolidating. Civil society is thinning. Antitrust enforcement is rebuilding from a low base. Encryption is contested. Whistleblowers face harder consequences. Federalism is being eroded in some areas (preemption, federal funding strings) and amplified in others. Elections themselves are contested. No single one of these is a five-alarm fire; the convergence is what political scientists worry about.

What can be done

Most readers don't write the laws or appoint the judges. The civic-action work is real and slow (Volume I, Lesson 24 covered this). What individuals can do:

— Maintain personal capacity for what you can\'t outsource — financial resilience, skill diversification, network depth.

— Use the checks that exist personally: encrypted communications, cash holdings, multi-jurisdiction relationships if your work allows.

— Support the institutions that work — independent journalism (paid subscriptions to outlets you trust), legal-aid and civil-liberties organizations, local civic associations.

— Engage with politics at the levels where engagement still has leverage — primaries, local elections, single-issue advocacy.

— Teach what you know. The wealth of historical and structural knowledge in the framework above is rare; sharing it widens the coalition.