The Department of Homeland Security has been shut down for 36 days. This is the third funding lapse in six months. More than 400 TSA officers have quit since February 14. On March 14, 55 percent of scheduled staff didn’t show up at Houston Hobby. Airlines are warning of chaos through Easter. On Saturday, President Trump announced that ICE agents would deploy to airports on Monday to help with security. On Sunday, border czar Tom Homan clarified that ICE would do crowd control and exit-guarding — not X-ray screening, which requires months of specialized training — while simultaneously continuing immigration enforcement operations. This morning, ICE is at the airports.
Both parties claim the other is holding the TSA hostage. Both parties are right. And that, precisely, is the problem.
Why the Obstruction Isn’t Working
A little over a decade ago I wrote a paper called “Signaling through Obstruction.”* The core argument: legislative obstruction — holding up a vote, blocking a nomination, withholding a funding bill — can convey genuine information about preference intensity, but only under specific conditions. The signal is credible when it is costly to the person sending it. A senator who refuses to fund an agency she controls, at real political cost to herself, is demonstrating — in the language of signaling theory — that her preferences on the underlying issue are genuine and strong. Cheap obstruction, by contrast, carries no information. If I can block something without paying any price, my blocking it tells you nothing about how much I actually care.
The DHS shutdown has a structural problem with this logic: the cost of obstruction is diffuse. Airport wait times — three hours in Atlanta on Sunday, lines past baggage claim — are a negative-valence good for everyone. They hurt TSA officers, obviously, who are working without pay and quitting at rates that will compound for months, since each new officer requires four to six months of training to certify.1 They hurt travelers. They hurt airlines — American, Delta, Southwest, and JetBlue CEOs signed a joint letter to Congress on Sunday. They will hurt the FIFA World Cup and the America 250 celebrations later this year.
What they don’t do, with any precision, is hurt the people causing them. Congressional members fly, of course — but they fly on schedules with security buffers, staff managing their movements, and the symbolic insulation of office. More importantly, each side has a workable account of why the other side is responsible for the pain. Democrats say Republicans won’t fund TSA separately from ICE. Republicans say Democrats won’t fund the full DHS. Both of these sentences are literally true. When both the obstructors and the obstructed can credibly claim to be victims, the obstruction has dissolved into noise. The signal of preference intensity is gone. What remains is a competition over who gets blamed — which is a different game entirely, with different equilibria.
This isn’t a failure of political will. It’s a structural feature. The signaling through obstruction mechanism requires a cost that falls specifically on the party that could end the impasse by conceding. Here, both parties can end the impasse by conceding, and both are paying costs from not conceding — but the costs fall primarily on people who aren’t at the table. That’s not an informative signal. That’s a coordination failure dressed up as a principled stand.
What ICE at Airports Actually Does
Trump’s Saturday Truth Social post said ICE would do “Security like no one has ever seen before, including the immediate arrest of all Illegal Immigrants who have come into our Country.” Homan’s Sunday clarification said ICE would guard exits and manage crowd flow “in areas that don’t need their specialized expertise” — specifically excluding X-ray screening — while continuing immigration enforcement operations on the side.
These are not the same job description. The gap between them is itself a signaling problem: which principal is ICE actually serving this week — the one who announced aggressive enforcement, or the one who walked it back to crowd control? Readers of this post will recognize the bilateral accountability structure: when an agent’s instructions from one principal contradict the stated expectations of another, and both are watching, the agent’s behavior on the ground is the resolution — not the announcement.
But there is a deeper institutional irony here, and it connects directly to a post we ran last month. Airports employ thousands of workers. A non-trivial share of those workers — baggage handlers, cleaning crews, food service staff, ground transport — are undocumented or have uncertain immigration status. The ICE presence at airports is not invisible to these workers. It is, in fact, precisely the kind of enforcement visibility that changes behavior: if showing up to work today means encountering ICE agents conducting simultaneous immigration enforcement operations, the calculus on showing up changes. The classifier reshapes the population it’s classifying. The deployment designed to address the TSA staffing shortage may, through this mechanism, accelerate it — not by removing TSA officers, but by reducing the auxiliary workforce that keeps airport operations functional.
This is not a hypothetical. The chilling effect on labor-force participation following visible immigration enforcement is one of the better-documented behavioral responses in the literature. Sending ICE to airports during a staffing crisis, in a context of heightened enforcement visibility, is not obviously a move that makes airports function better. It is, however, a move that makes a particular audience — one watching Truth Social — feel that something decisive is happening.
The Reforms They Were Asking For
Democrats withheld DHS funding specifically to force reforms to ICE enforcement practices: judicial warrants for home and business entries, removal of masks, visible identification, limits on enforcement at sensitive locations. These demands were triggered by the January killings of Alex Pretti and Renee Good in Minneapolis — two US citizens killed by DHS agents during immigration enforcement operations. The reforms are targeted at specific practices in specific contexts: masked agents entering private properties without judicial warrants.
Now consider what ICE agents at airports actually encounter. Airports are public spaces. The warrant requirement — the centerpiece of Democratic demands — has no purchase there. ICE can conduct enforcement operations in airports without a warrant because airports are not homes or businesses. The mask question and visible ID question are theoretically live, but in an airport environment where ICE agents are nominally assisting with security, the norms governing their appearance are genuinely ambiguous. The three specific reforms Democrats demanded don’t structurally apply to airport enforcement.
The conservation of impossibility applies here: the structural problem Democrats were trying to fix doesn’t disappear when you change the enforcement venue. It relocates. ICE in airports isn’t operating under the constraints the Democrats were trying to impose — not because the administration defied those constraints, but because the geographic relocation puts the enforcement outside their scope. The impossible problem migrated from Minneapolis-style home raids to airport enforcement operations, wearing a different face.
Whether this is the administration’s intent is, as usual, a question the model can’t answer from the outside. What the model can say is that the observable signature of this move — enforcement that looks aggressive to one audience, helpfully logistical to another, and structurally immune to the specific reforms being demanded — is consistent with a strategy that is simultaneously placating the Truth Social audience, answering the Homan-clarification audience, and operating in a legal gray zone the Democratic demands don’t reach.
Thirty-six days in, the shutdown has produced this: TSA workers are leaving the profession at rates that will take six months of training to replace. Airport chaos is peaking during spring break and will worsen. Both parties have successfully narrated the other as the obstructor. ICE is at the airports, doing something — the exact nature of which depends on who is describing it and to whom. And the specific reforms Democrats were signaling for, through 36 days of painful and expensive obstruction, are not the reforms that airport-based ICE enforcement is subject to.
The signal, in the end, was costly. It just wasn’t informative.
With that, I leave you with this.2
1 The attrition problem compounds in a specific way that deserves emphasis. A TSA officer who quits today cannot be replaced for four to six months. Each new officer requires certification training at the TSA Academy. The shutdown is not just creating a current staffing shortage — it is destroying future staffing capacity. Even after the shutdown ends and back pay arrives, the workforce will be smaller and the pipeline to replace it will be slow. The costs of the obstruction are being paid not just by today’s travelers but by travelers six months from now.
* A note for formally-inclined readers on why this case is structurally harder than the paper it’s linked to. “Signaling through Obstruction” analyzes unilateral obstruction: one party bears a cost to signal preference intensity to a principal who cannot directly observe it. That case has a well-defined equilibrium — the threshold the obstructor must meet is set by the principal’s prior beliefs and updated against the observable cost of obstruction. The bilateral case here is different in a specific way. When both parties are simultaneously obstructing to prove relative commitment, the threshold each must meet is endogenous — it depends on what the other party is doing. No fixed finish line exists to cross.
Game theorists have a name for this. A war of attrition is a game in which two players compete by simultaneously bearing costs over time, and the winner is whoever can last longer. The classic examples are animal contests (two stags locking antlers, each paying a cost per unit time), patent races, and — less glamorously — funding standoffs in a legislature. The key property of war-of-attrition games is that, in the mixed-strategy equilibrium, each player randomizes over when to concede — meaning the expected duration of the contest, and the total costs borne, can be arbitrarily large. There is no natural stopping point. The game ends when someone quits, and the incentive to not be the one who quits is precisely what keeps both players locked in. Escalation is individually rational; collective ruin is the result.
This is the tragedy of the commons in slightly different clothes. In the standard commons problem, individually rational extraction by concurrent agents destroys a shared resource neither wants destroyed. Here it is concurrent withholding — each party rationally refusing to concede — that destroys the shared resource: a functioning airport security system, a trained TSA workforce, and (not trivially) public confidence that the government can perform basic operational tasks. Both parties want airports to work. Both parties want to be seen as the one who cared more about making them work. The Nash equilibrium of their bilateral signaling game is producing the outcome neither actually wants. This is not a novel result. It is, however, a useful reminder that “both sides are being rational” and “the outcome is good” are not the same sentence.
2 Stealers Wheel, “Stuck in the Middle with You.” Clowns to the left, jokers to the right. The TSA officers, the travelers, and the airports didn’t write the song. They’re just playing it.
1 thought on “Stressing at the Airport? Just chill…ICE is on the way!”
Comments are closed.