The Midterm as a Hostile Takeover

America is approaching the midterms with two competing theories of election management. One treats voting as a decentralized public function: tedious, state-run, legally bounded, and therefore rather resistant to anyone’s personal sense of destiny. The other treats it as an underperforming subsidiary awaiting decisive federal intervention.

That second theory is no longer confined to the more excitable corners of the commentariat. An outside Trump-linked adviser, Peter Ticktin, told The Atlantic that possible outcomes included Trump trying to declare martial law around November 3; Ticktin’s formulation was: “Either martial law, or step aside,” as reported here. The article does not establish that Trump has decided on such a course. It does establish that an adviser with longstanding ties to him is saying the quiet part into a microphone.

The relevant constitutional nuisance is that elections are principally administered by states and localities. Yet Trump has called for nationalizing voting in at least 15 places, and amplified a proposal for a national-security emergency imposing citizenship-proof and voter-ID requirements. The political move is more important than the legal flourish: recast ordinary election rules as a security crisis, then imply that exceptional executive machinery is merely prudent housekeeping. Every would-be centralizer eventually discovers that paperwork looks much grander when illuminated by emergency lighting.

This is not simply a dispute about fraud claims, which the reporting describes as unsupported or false in key instances. It is a struggle over who may define the conditions of political legitimacy before votes are counted. That is why the map fights, the registration-system failures and the rhetoric of emergency belong in one issue. They are all arguments over control of the electoral operating system.

In Missouri, the fight is less cinematic than martial law and more immediately operational: which congressional map governs an election in which absentee voting has already begun? The state’s Trump-backed map, designed to help Republicans gain a House seat, has returned to the Supreme Court for the third time after a federal appeals court sought to revive it at the last minute according to the Associated Press.

The Supreme Court has twice blocked use of the new map for November. Missouri’s own high court held that the districts could not take effect without a statewide referendum, leaving the previous map in place. That ought to settle the immediate administrative question, at least in the antique constitutional order where rules are known before ballots are printed. Instead, more than a million potential voters face the possibility of being shifted among districts while voting is underway.

The map’s purpose is not coy. Republicans hope it will help them win seven of Missouri’s eight House seats, rather than six, including through a substantial reshaping of the Kansas City-based district represented by Democrat Emanuel Cleaver. Mid-decade redistricting, encouraged by Trump in Republican-led states, is the electoral equivalent of rearranging the chessboard after noticing one’s opening was not quite as brilliant as advertised.

There is a distinction worth preserving amid the procedural fog. Legislatures may pursue partisan advantage; courts may review whether they followed the governing law; voters may be entitled to a referendum. But an election system cannot function as a rolling injunction docket. The point of administration is that citizens should know not just how to vote, but what political community their vote is being counted within.

Texas Republicans drew new districts using the most seductive data point in politics: last time. Trump’s unusually strong 2024 performance among Latino voters helped justify a map that made several seats friendlier to the GOP while retaining large Latino populations. In the reconfigured 15th District, Latino residents comprise 74.5% of the citizen voting-age population, and the district would have given Trump an 18-point win in 2024 Reuters reports.

But district lines are a storage device for past coalitions, not a warranty on future ones. Reuters found some South Texas Latino voters who backed Trump expressing frustration over economic pressure and the administration’s immigration crackdown. A Cook Political/TelevisaUnivision poll of 1,000 Texas Latinos found that 18% of those who voted for Trump no longer supported him, while almost two-thirds said they would be less likely to vote for candidates backing Trump’s immigration policies. These are signals, not prophecy; Republicans reject the idea that Latino voters are departing.

The 15th is consequently a useful test case. Republican incumbent Monica De La Cruz faces Democrat Bobby Pulido, a singer, and an August Texas Southern University poll had Pulido ahead by five points. Forecasting outfits differ on the race: Sabato’s Crystal Ball calls it a toss-up, while Cook Political rates it Lean Republican. The disagreement is not a flaw in democracy. It is what democracy looks like before consultants have converted uncertainty into a donor email.

The deeper contradiction is brutally ordinary. A party can construct a district around a presidential result, but it cannot redraw the price at the pump or make a harsh immigration policy feel less harsh to families living with its consequences. In McAllen, average gasoline was $4.09 a gallon as of September 22. The electorate has a persistent habit of treating abstractions about coalition realignment as secondary to the cost of moving through the week.

The other great temptation is to turn administrative failure into a theory of national sabotage. Reuters found that states may have mistakenly added more than 30,000 self-declared non-citizens to voter rolls since 2000, largely through software glitches and clerical or processing errors connected to license and state-ID applications in this report. That is a real administrative problem. It is not, on the available evidence, an organized scheme to register immigrants illegally.

The details matter because they puncture the preferred mythology of both complacency and panic. New Jersey said a technology glitch added about 6,600 non-citizens between June 2023 and June 2024 despite their answering no to a citizenship question. Oregon’s mistakes included immigrants using foreign documents and American Samoans registered without officials recognizing that U.S. nationality is not the same legal category as citizenship. Iowa’s initial estimate above 2,000 narrowed to 277 after six months; officials said 35 non-citizens cast counted ballots in a 2024 election with more than 1.6 million votes.

Reuters could not determine how many mistakenly registered people nationwide actually voted. Its review also found only 129 federal prosecutions under the relevant statute since 1996, with no alleged coordinated effort to influence an election. Experts cited in the report said the numbers were too small relative to more than 170 million registered voters to affect outcomes. This is what boring reality looks like: databases fail, forms are misprocessed, initial estimates are revised, and the resulting problem requires competent repair rather than a constitutional bonfire.

That is the midterm’s actual dividing line. Systems need accurate rolls, stable maps, transparent rules and remedies proportionate to the evidence. They do not need a president hovering over state election offices like a private-equity owner who has mistaken the county clerk for a hostile acquisition. The most durable election security is not theatrical force. It is public administration good enough that fewer people can plausibly pretend force is necessary.