The Selective State

The Trump administration’s governing theory increasingly looks less like small government than selective state capacity: the state is vigorous where it can punish, investigate, demand records, or confer permission, and oddly frail where it must quietly coordinate expertise before something breaks.

That distinction matters more than the day’s personality drama. Cybersecurity, fire response, visa power and family policy may seem like separate rooms in the federal mansion. They are, however, connected by the same renovation plan: cut the unglamorous systems that make collective life function, then retain—or enlarge—the discretionary levers that put individual people at government’s mercy.

The result is a peculiar constitutional mood. Washington is not vanishing. It is becoming more personal, more politicized, and less reliably competent at the boring work that citizens notice only when the water stops running, the fire crosses the road, or the election website goes dark. That is not an argument for bureaucracy as a sacred object. It is an argument for recognizing that capacity is power, and that power is being redistributed rather than retired.

The most ominous vulnerability in the latest reporting on CISA is not that a foreign government will secretly alter millions of paper ballots. The article says that risk remains low in a decentralized system with largely offline voting equipment. The more plausible objective is cheaper and more modern: disrupt registration systems or results websites, generate confusion, and let domestic actors do the confidence-destruction for free.

This is the environment in which CISA reportedly lost a third of its staff, saw its budget reduced, and has operated without a permanent Senate-confirmed director since Trump returned to office. Federal election-intelligence briefings for state officials were halted, and funding for an information-sharing network ended, according to The Atlantic. After suspected Iran-linked hackers targeted water and wastewater systems in more than a dozen states, Arizona’s election office reportedly had state homeland-security officials contact CISA without identifying the affected office. That is not a healthy federal partnership. It is a witness-protection program for intergovernmental trust.

The irony is almost too neat: an agency created during Trump’s first term became suspect after it defended the integrity of the 2020 election against Trump’s false claims. Now the administration is reportedly investigating 2020 elections, seeking voter data, and threatening recalcitrant election officials, while state officials must prepare for 2026 disruption without the federal threat analysis no contractor consortium can replicate. One can reduce an agency, of course. One cannot reduce the adversaries’ interest in exploiting the resulting gap.

California’s Timber Fire was burning through Big Sur with only 12 percent containment in the reported Saturday update, threatening 503 structures and prompting closures and evacuation orders along Highway 1. Near Reno, the Bug Fire had consumed roughly 93,381 acres and displaced at least 4,000 people earlier in the week. These are not abstractions about resilience. They are personnel counts, engine assignments, helicopter availability, wind, heat, dry fuel, closed parks, interrupted routes, and people removed from care facilities.

Yet the federal workforce behind this response has been placed under another version of the same capacity test. The National Federation of Federal Employees says budget cuts and layoffs under Trump, including DOGE-driven reductions, have deepened uncertainty for federal wildland firefighters. Its claim that more than 6,000 Forest Service workers were cut—thousands qualified for wildland fire duty—has not been independently substantiated in the supplied report, and the administration’s response is not included. Still, the structural problem predates any particular head count: firefighting has long leaned on workers facing poor pay, weak benefits and punishing living costs.

A government can call this streamlining, but fire does not honor the euphemism. It merely asks whether the crews, research, equipment and local knowledge exist when a bad afternoon becomes a bad week. The administration’s broader lesson is that cuts are easiest to announce in a spreadsheet and hardest to defend from a road closure. The public service being trimmed is rarely named “the thing that saves your town.” It is named after an office, a grade, or a line item—which is rather the point.

The reported revocation of Andrés Manuel López Beltrán’s U.S. visa offers the inverse picture of federal power: no shortage of discretion, and very little public explanation. López Beltrán, son of former Mexican president Andrés Manuel López Obrador, says the action was politically motivated and that no evidence of wrongdoing was given. The embassy declined comment; the State Department invoked the confidentiality of individual visa cases. Mexico’s president, Claudia Sheinbaum, says she knows of no Mexican prosecutorial investigation into him.

Those facts do not establish that the revocation was improper; visa decisions can involve information that governments cannot responsibly disclose. Nor do they establish the basis for this particular action. But they expose a democratic difficulty built into the instrument. Reuters reported that the administration had revoked visas for at least 50 Mexican politicians and officials across the political spectrum amid its cartel crackdown. When entry to the United States can be withdrawn without a public case file, the measure can be a legitimate national-security tool—and an extraordinarily potent political signal. Often, it can be both from the observer’s point of view.

This is where the newsletter’s two versions of state capacity meet. Election officials are reportedly left seeking substitutes for federal intelligence; firefighters face an uncertain federal workforce; but a visa can still operate as a cross-border judgment with immediate personal and electoral consequences. The choice is not between a powerful state and a weak one. It is between a state that invests in public capabilities and one that favors individualized coercive capabilities. The latter is more dramatic. The former is what keeps a republic from having to discover its infrastructure by emergency.