Nobody's leaving now
The same mechanism, running backwards. Plus the bit I got wrong when I wrote about this in 2021.
I wrote a post in the summer of 2021 called “Everyone I know is leaving.” Three people had gone in eight weeks from the agency I was at then, the market was extraordinary, and I worked through why staying somewhere is structurally a real-terms pay cut when salaries are rising.
I’ve had one recruiter message this month. Two of those three people have been made redundant since Christmas, both from companies that were hiring aggressively when they joined.
So it seems worth going back to that post rather than leaving it sitting there as a confident description of a world that lasted about a year.
The same mechanism, backwards
The thing I described is real. Internal pay is anchored to what you were paid last year; external offers are anchored to what the role costs today. That’s a genuine structural feature of how compensation works and I still think it’s the single most useful thing to understand about your own salary.
What I didn’t say, because I hadn’t thought it through, is that it runs in both directions.
In a falling market the anchoring works the other way. The person who stayed is now anchored to a number set when the market was hot, which means they’re being paid above what the same role would cost to fill today. That’s pleasant for a while, and it also makes you the expensive line on a spreadsheet when somebody is choosing where to cut.
And the person who moved for a big increase 18 months ago is now the most recently hired, on the highest salary, with the least tenure, at a company that over-hired. Which is not a comment on their judgment. It was the right decision with the information available. It’s just that “the right decision” and “the good outcome” are different things, and a market can put a lot of distance between them.
The right decision and the good outcome are different things, and a market can put a great deal of distance between them.
What I got wrong
Not the mechanism. The tense.
I wrote about the loyalty penalty as though it were a permanent property of employment, and it’s a permanent property of rising markets, which is a different claim. Everything in that post is contingent on a condition I hadn’t noticed I was assuming, because it had been true for the entire period I was looking at.
Which is the most ordinary analytical error there is: taking the conditions of the last few years for the way things are. I’ve done it about tooling, I’ve done it about browsers, and apparently I do it about labor markets too.
The honest version of that post would have said: in a market like this one, staying costs you, and markets change, and when this one turns the arithmetic will invert and so will all the advice.
What it’s actually like
I want to be careful not to write about this as though it were only an interesting mechanism, because it’s happening to people.
Somebody I know had 60 days to find a new job or leave the country, because their right to be here was attached to the employer that let them go. 60 days, in a market where the process alone takes longer than that.
Somebody else moved cities for a role in 2022, signed a tenancy, and was cut in the first round because last in is the cheapest to remove.
And the ones who stayed are doing the work of the ones who went, which is exactly the cycle I described from the other end in 2021: fewer people, same clients, same deadlines, and the absorption is invisible because everyone is competent and doesn’t complain immediately.
Agencies aren’t outside this, incidentally, and I’m in no position to sound smug about it. I left three months after writing that post, spent a year in-house, and came back to agency work last spring, which is close to the least stable 18 months a person could have picked. Marketing spend is among the first things a business cuts, so agency work follows a downturn with a lag of a few months. We’re all right at the moment. “At the moment” is doing some work in that sentence.
We keep calling a cycle a truth
The thing I’d most like to hold onto from all of this.
In 2021 the consensus was that the balance of power had permanently shifted toward workers, that remote was irreversible, that talent scarcity was structural. All of that was published confidently, at length, including a version by me.
Right now the consensus is that the era of cheap money is over for good, that the industry over-hired for a decade, that everything is correcting to a new permanent level. Also published confidently, at length, often by the same people.
Those are the same error made twice, 18 months apart, in opposite directions. Each time the present conditions get described as the end of a process rather than a point in one, and each time it’s more compelling than the boring answer, which is that this is cyclical and neither of the confident versions will age well.
I don’t know what happens next and I’d be suspicious of me if I did.
What I’d actually say now
Three things, hedged, and I’d apply all of them to myself.
Have a runway. Whatever you can manage. It’s the only thing on this list that converts from a feeling into an option, and it’s the same point I was making about saying no in 2021: your capacity to make good decisions is mostly a function of how much room you have.
Keep the relationships that aren’t your employer’s. The people who found work quickest this year were the ones with a network that predated the job, and that network is much easier to maintain in good times than to build in bad ones.
And the question from 2021 still stands, it’s just got a different answer. Are you staying because you’ve chosen this, or because you haven’t looked? That was worth asking when leaving was easy. It’s worth asking now, when it isn’t, because if the answer is “I’ve chosen this,” that’s a genuinely good position to be in and worth knowing.
Anyway, if you’re in the middle of it, I’m sorry, and it’s not about you, and I hope it’s short 😔