The price of leverage

There’s a coworking space in Condesa where I do most of my thinking, the kind where the espresso machine sounds like it’s in pain and the wifi drops every time someone joins a video call. I was sitting there Tuesday morning, dry-season haze turning the windows a particular shade of beige, when I read that Delta Air Lines had suspended Congressional travel perks until Congress funded the TSA.

I read it twice.

Then I posted about it, and apparently I wasn’t the only one who found it clarifying in a way that felt almost obscene.

Delta did in one gesture what two decades of union organizing couldn’t. TSA workers had been calling in sick, walking off, leaving airports to buckle under the weight of a shutdown that left them working without paychecks. Their unions had been asking for real leverage, the kind that makes powerful people feel the inconvenience of a problem, for years. They got nothing. Delta, protecting its revenue, protecting its operational continuity, protecting its brand from the particular hell of a terminal full of stranded passengers, made one decision. Congress felt it immediately.

The Senate approved TSA funding by Friday, March 28th. Five days. That’s the timeline.

I want to be fair to the mechanics here, because the counterargument is real: Delta wasn’t being heroic. CEO Ed Bastian had already been publicly calling for a resolution before the perks got yanked. The move was economic self-interest, not solidarity. Delays cost airlines money. Stranded passengers become former passengers. The calculation was ruthless and boring and had nothing to do with the workers.

That’s exactly the point.

The workers’ leverage was moral. Their argument was correct. They were being asked to staff critical federal infrastructure without pay, which is a thing that shouldn’t happen in a country that calls itself functional. They had every reason to demand better. They had two decades of reasons. None of it moved the needle the way one airline’s inconvenience calculation did, in one week, in March 2026.

What Delta had that the unions didn’t was structural. It had something Congress actually used and didn’t want to lose. That’s the whole game. Leverage isn’t about being right. It’s about controlling something the person across the table needs more than they need to ignore you.

the unpaid math

I’ve been thinking about this alongside something I posted a couple days later, about what some researchers call the “race tax.” Black professionals spending somewhere between two and seven extra hours per week on DEI work, mentorship pipelines, visibility labor, the kind of institutional maintenance that keeps diversity numbers looking presentable in annual reports. Their white peers skip it almost entirely, not through malice necessarily, just through the ordinary privilege of not being asked. Not being needed for that particular function. The math comes out to somewhere between 100 and 350 unpaid hours a year, depending on the role, the company, the moment in the news cycle.

Nobody puts that on the offer letter.

What strikes me about this alongside the Delta thing is the same structure underneath both. In one case, workers who staff federal security infrastructure have their leverage systematically defused for twenty years. In the other, a specific subset of workers perform an entire category of institutional labor that never shows up in their compensation, their title, or their performance review (unless it goes wrong, in which case somehow it does). Both are stories about work that the system requires but refuses to pay for. Both are stories about who gets to hold leverage and who gets to be leveraged.

The person who can make Congress feel an airport delay gets funded in five days. The person who spends seven hours a week keeping their company’s DEI numbers viable gets a LinkedIn recommendation and, if they’re lucky, a “culture contribution” mentioned in their annual review.

Negotiate accordingly, I said in the post. I meant it. Though I’m aware that “negotiate accordingly” is advice with a ceiling, which is that you can only negotiate for what the other party is willing to acknowledge exists.

the moat

The third piece of this, the one that got the least traction but that I keep returning to, is about AI and tenure.

There’s a narrative running through tech right now about automation as this equalizing force, this great reshuffling that will affect everyone equally, that no job is safe, that we’re all in this together. It’s a very useful narrative if you’re the one doing the automating.

The reality, as far as I can tell, is more specific and more vicious. The people who can automate their jobs without consequence are the people who already have job security: the senior engineers, the tenured faculty, the directors with fifteen years of institutional knowledge who are interesting to keep around even as their immediate function gets absorbed by a model. They get to be the ones who “work with AI,” who “oversee the process,” who transition into new roles that the company invents for them because they’re worth retaining.

The people who never got hired because “budget” or “headcount freeze” or “we’re being strategic about growth right now” are watching from outside as those same roles get replaced. They weren’t in the building when the music stopped. They don’t get a chair.

This isn’t a coincidence. The budget freezes that kept entry-level and mid-level roles unfilled over the last few years created the exact conditions where automation looks like a clean solution rather than a displacement. If you never hired the person, you never have to fire them. The math was always the point. I said that in the post and I still think it’s true.

What connects all three of these things, the TSA funding fight, the race tax, the AI moat, is something about who gets to have leverage treated as legitimate. Delta’s leverage was legitimate because it was structural and it cost people money. Union leverage was illegitimate for twenty years because it was moral and it only cost people dignity. The unpaid DEI labor is illegitimate as a negotiating chip because acknowledging it would require companies to either pay for it or admit they’ve been extracting it. The tenure moat is legitimate because the people inside it have already proven their value to the people making the decisions.

There’s a version of this essay that ends with something about organizing, about building the kind of structural leverage that makes the powerful feel it. That version is true and I believe it. But I’m sitting in Condesa in the dry-season haze and I keep coming back to something simpler and uglier: we already know what works. We’ve watched it work, in five days, in March 2026. The question was never how. The question was always who gets to try it.