When we last checked in on the DHS shutdown, the diagnosis was a bilateral war of attrition: two parties simultaneously bearing costs, each hoping the other would flinch first, while the costs fell primarily on TSA workers and travelers who had no seat at the table. The shutdown was 36 days old. The signal, I argued, was costly but not informative — both sides could credibly blame the other, which dissolved the signaling mechanism that makes obstruction useful.
That was last Monday. Here is where things stand today.
The shutdown is now 44 days old — the longest in US history, surpassing the record 43-day shutdown last fall. On Friday, the Senate passed a bill funding DHS except ICE and parts of Customs and Border Protection, in the early hours of the morning. The House woke up, called it a joke, and passed its own 60-day continuing resolution funding all of DHS — including ICE — by a 213-203 vote, with three Democrats from competitive districts crossing over. Both chambers then left town for a two-week recess. The Senate’s bill is dead on arrival in the House; the House’s bill is dead on arrival in the Senate. Congress returns April 13–14. Nobody expects resolution before then.
And yet: on Friday, President Trump signed a presidential memorandum directing the Department of Homeland Security to pay TSA workers using funds with a “reasonable and logical nexus to TSA operations,” citing an emergency situation compromising national security. The TSA, which has been operating without pay for 44 days, will begin receiving paychecks as early as today.
The shutdown is still ongoing. The workers are getting paid. I want to explain why that combination is formally strange — and why it matters for the next shutdown, and the one after that.
Paying the Hostage
The standard hostage-taking model in legislative bargaining works like this. Party A holds something Party B values. Party B’s willingness to meet Party A’s demands depends on how much it values the thing being held, and how much it expects to gain by conceding. The hostage’s suffering is load-bearing: it is what makes the threat credible and what puts pressure on Party B to act. If you remove the suffering without the concession, you don’t resolve the standoff — you just change what the standoff is costing each side.
In the previous post, I described the DHS shutdown as a game where the “hostage” — TSA workers without pay, airport chaos, travelers in four-hour lines — was nominally held by both parties simultaneously, which is why the signaling mechanism had broken down. Both sides could claim to be the one who wanted to free the hostage. Neither could credibly claim the other was the sole captor.
Trump’s memo introduces a third move nobody modeled: the executive branch reaching into the standoff and unilaterally paying the hostage, without congressional authorization, while the standoff continues. The workers get their checks. The ICE dispute is unchanged. Congress is on recess.
In one sense, this is simply good politics: airport chaos was the most visible cost of the shutdown, and visible costs during recess are dangerous for the party in power. By paying TSA workers, the administration reduces the most salient public symptom of the impasse without conceding anything substantive. Democrats lose their best visual — unpaid workers, closing airports, the Easter travel weekend — and Republicans go home to their districts with something to point to. The underlying dispute about ICE reform is exactly where it was on February 14.
But there is a structural consequence that outlasts this particular standoff, and it cuts in a direction that should concern anyone who cares about congressional leverage in future funding fights — regardless of which party they’re rooting for.
What the Memo Does to Future Shutdowns
The shutdown as a legislative tool works because shutdowns are painful. The pain is the mechanism. A legislature that withholds funding from an executive agency is, in effect, threatening to impose costs on the public unless the executive capitulates on some policy demand — or vice versa. The threat’s credibility depends on the executive’s inability to relieve those costs unilaterally. If the executive can always pay the workers, open the parks, restore the services — whenever the political cost of not doing so exceeds some threshold — then the legislature’s funding lever is degraded. Not broken, but meaningfully weaker.
Notice the asymmetry this creates. The executive can selectively relieve shutdown costs when they are politically inconvenient — right before a recess, right before a holiday travel surge, right before an election — and allow costs to accumulate when they are generating pressure on the opposing party. This is not a hypothetical abuse; it is what the TSA memo did. Congress couldn’t agree on how to fund DHS. The administration unilaterally paid one of the most visible and sympathetic groups of federal workers right before a two-week recess, using emergency authority, citing national security, and daring anyone to sue.
Whether anyone will sue — and whether such a suit would succeed — is a genuine open question. The presidential memo directs Mullin and OMB director Russ Vought to use funds with “a reasonable and logical nexus to TSA operations.” Whether that instruction has a legal basis is contested. What is not contested is the structure: the executive branch spending money that Congress has not appropriated, on the grounds that the failure to appropriate constitutes an emergency.
Readers of this blog will recognize the family resemblance. IEEPA gave the executive essentially unchecked authority to impose tariffs, on the grounds that a declared emergency authorized sweeping economic powers. The Supreme Court ruled in February that this exceeded statutory authority. The TSA memo is not IEEPA — it’s a narrower instrument, with a different legal basis, and the political context is different in ways that matter. But the underlying move is the same: executive emergency authority deployed as a substitute for legislative action that Congress will not take.1 The Court slapped that down in one domain six weeks ago. Nobody is talking about the parallel in this one.
The Game When Congress Returns
When the House and Senate return the week of April 13, the formal structure of the negotiation will look like this. There are two bills that have each passed one chamber and are dead in the other. There is a two-week recess separating today’s impasse from the next attempt at resolution. And there is now a precedent — however contested its legality — that the executive can relieve shutdown pressure on a timeline of its own choosing.
That last element changes the Democrats’ leverage in a specific and underappreciated way. The war of attrition in a shutdown works because both parties are accumulating costs over time and hoping the other side’s cost tolerance runs out first. Democrats’ cost tolerance in this standoff has been sustained, in part, by the public visibility of airport chaos — the daily images of four-hour security lines, the airlines’ joint letter to Congress, the TSA officer quitting rates.2 That visibility was their evidence that the standoff was imposing real costs on real people, and that Republicans were responsible for refusing to end it.
The memo does not end the shutdown. But it reduces the visibility of the shutdown’s costs, on a timeline the administration controls, going into a recess in which members will be back in their districts answering questions about it. When the Senate convenes again on April 13, the airport crisis will be two weeks old and somewhat resolved. The ICE reform demand will be exactly as unresolved as it is today. The question is whether Democratic senators in competitive states will conclude that the diminished visibility of the shutdown’s costs — combined with the continued pressure of the May 22 House CR deadline — makes concession more attractive than it was on Friday morning.
This is conservation of impossibility, but applied to leverage rather than to a formal impossibility result. The underlying 60-vote problem in the Senate hasn’t changed. The ICE reform dispute hasn’t changed. What changed is the distribution of visible costs across the two parties — and that distribution was changed not by a legislative agreement but by executive action that may or may not survive legal scrutiny. The impossible problem didn’t disappear. It was temporarily covered up, on one side’s timeline, with emergency authority of uncertain legality.
That is not a resolution. It is a decision about whose costs are visible this week. And those are not the same thing.
TACO
Over the weekend, Trump announced he would refuse to sign any legislation — any legislation — until the SAVE Act lands on his desk. The SAVE Act is a voter ID bill for federal elections that has passed the House but faces long odds in the Senate. The threat, taken literally, means Trump would withhold his signature from must-pass legislation, reconciliation bills, and anything else in the queue until the Senate produces a bill that currently doesn’t have 60 votes.
In the context of a war of attrition, this kind of announcement is a commitment device: a costly signal intended to demonstrate that your cost tolerance is higher than the other side believes. The problem is that commitment devices require credibility, and credibility requires a track record of actually holding the line.
Trump signed the TSA memo on Friday. The workers get paid Monday. Congress is on recess. No concession was extracted.
In the same 48-hour window, Trump unilaterally relieved the most politically visible cost of the shutdown — without a deal, without a concession, before Congress even returned — and then announced that he would hold firm on everything else until the Senate produces a bill it currently cannot pass. House Republicans who spent six weeks absorbing the political cost of his ICE funding demands, including presiding over the longest shutdown in American history, are now being asked to believe that the next ultimatum is different. They are being asked this by the same principal who just demonstrated he’ll pay the hostage before the negotiation concludes.
There is a term, coined during the tariff fights of 2025, for the phenomenon of Trump issuing a maximalist threat and then quietly retreating before it bites: TACO, for Trump Always Chickens Out. The label is uncharitable, but the formal observation underneath it is sound. In the Kalshi post, I quoted Kenny Rogers: know when to hold ’em, know when to fold ’em, know when to walk away. The problem isn’t folding — folding is sometimes correct. The problem is folding and then announcing, with a straight face, that you’ve never been more committed to holding.3
House Republicans have been at the table for this hand. They’ve seen the cards. The question for April 13 isn’t whether Democrats will blink. It’s whether the Republicans who took the beating have any reason to believe the next all-in is real.
With that, I leave you with this.4
1 The legal basis the memo cites — funds with a “reasonable and logical nexus to TSA operations” — is doing a lot of work. TSA is funded through the DHS appropriation that lapsed on February 14. The memo is, in effect, asking OMB to identify existing funds in a different bucket and redirect them toward TSA payroll on the theory that TSA operations are nationally important enough to constitute an emergency. This is not obviously wrong as a matter of executive power doctrine; emergency reprogramming of funds is a recognized executive tool. What makes it structurally interesting is the context: this is not an emergency that arose externally. It is an emergency that arose because Congress chose not to appropriate the funds. The executive is using emergency authority to address a situation the legislature created by withholding ordinary authority. Whether courts will treat that distinction as legally meaningful is, at minimum, a question worth asking.
2 The attrition problem I flagged last week has continued to compound. Nearly 500 TSA officers have now quit during the shutdown. Each one takes four to six months to replace through the certification pipeline. The memo pays current workers, but it does nothing about the workers who already left. The structural damage to TSA staffing capacity is baked in regardless of when the shutdown ends.
3 To be precise about what TACO captures and what it doesn’t: the label describes a pattern of announced commitments that are subsequently abandoned, which in the language of game theory is a failure of credibility rather than a failure of strategy. Folding at the right moment can be optimal. What TACO names is the combination of folding and then claiming you haven’t — issuing a new commitment on top of the revealed one, as if the audience didn’t notice. In signaling theory, this is worse than simply having low cost tolerance. It is having low cost tolerance while burning the signaling channel you’d need to communicate high cost tolerance in the future.
4 Kenny Rogers, “The Gambler.” We used this one recently, and I considered a different closer. But it is the correct song, and I refuse to use a worse song because I already used the correct song. Know when to hold ’em.
3 thoughts on “Paying the Hostage”
Comments are closed.