Sometimes there’s a story circulating in the zeitgeist and I have to write about it. Today I’m actually publishing those thoughts instead of shelving them.
There’s a story going around the tech-world (and the people who invest in the tech-world) that laying off staff right now is important, necessary, and a sign of good judgment. It goes hand-in-hand with messages about having over-hired during the pandemic, and messages about an expected recession. I suspect it’s misguided, even if the other connected messages are true and accurate.
Honestly, this current layoff-streak looks to me like companies chasing each other, trying to convince investors that they’re serious and diligent. Better yet, that they’re responsible.
Turns out James Surowiecki sees the same pattern. He has an article on this here. He thinks this looks like a parallel to stories about “downsizing” from decades ago. I see any number of connections to other stories about austerity and responsibility.
I might judge this differently if it seemed like the tech companies involved were firing according to a plan or a consistent basis for judgement. Instead, it looks like they’re firing willy-nilly at all levels. Beyond that Business Insider article, the reports I’ve heard from people working in the affected companies all suggest that this current wave of layoffs is haphazard at best: one reported that the company was still hiring, often for the very positions they just opened by firing experienced workers. It seemed their company paid little attention to the composition of the teams they were firing from, or the relevant experience of the people they fired.
All of that makes me think that this move, firing roughly 6% of a company’s workforce, is more about needing to be seen as doing something rather than about doing the right thing.
So what are the costs incurred by this? If these companies are firing at scale, firing 6% of their employees while still making profit and while still sitting on war chests of as-yet uninvested funds (e.g. Ruth Porat’s segment of Google’s Q4 2022 call)—is that actually a sign of good judgment? Or are these companies shooting themselves in the foot?
In other words, is that story about firing people being the diligent, serious thing to do either true or accurate?
For argument’s sake, I’m going to accept as given that the other two stories are true. I’ll agree that these companies over-hired during the pandemic, and that the economy will either experience a recession this year or at least perform worse than hoped.
I’m not going to dispute any of that.
But what price do these companies pay with these firings? Does it make sense to fire these workers?
Let’s step back. When does it make sense to fire someone?
I posit two general cases:
1) It makes sense to fire someone when they hurt a company more by staying than they do by leaving.
2) It could makes sense to fire someone if their role is no longer relevant to the tasks performed by a company—but that ignores the benefit of reassigning them to a different role and saving on onboarding costs.
I admit, there may be other cases. But I think these two cover the majority of reasonable firing situations.
#2 is probably easiest to understand as a company-wide refocusing of effort. If a company totally shuts down an operation, stops trying to make or support a certain product, etc., it might make sense that some of the staff involved would no longer be relevant to the company’s goals. In that case, firing could make sense.
In counterpoint, I maintain that the company is probably better served by retaining whatever staff it can and assigning them new work in whatever the company’s area of focus may be. Onboarding staff takes time. Training staff takes time. Even if you have to retrain reassigned staff, you’re probably still saving money by conserving your staff pool.
Also, I haven’t seen a dramatic shift in stated plans among the various tech companies that are so eager to fire people. Thus, I don’t think these layoffs fall into case #2.
As for #1… what does it take for an employee to hurt a company more by staying than by leaving?
I’m going to ignore the case of an employee being a toxic piece of shit who harms others around them. That seems like a very clear cut entry in case #1, and it doesn’t match what I’ve heard about the current wave of tech layoffs. If that’s the reason for these firings, they’re firing a lot of the wrong people.
So. How does one calculate the costs of someone staying? How does one calculate the costs of someone leaving?
The costs of keeping an employee can be loosely summed up as “pay, benefits, and overhead.” Employees cost salary or wage, they cost payroll tax, they cost whatever is offered as benefits… there are probably other things I’m missing here: I’m not an expert and this is not intended to be professionally rigorous. But a lot of the costs associated with employee benefits and overhead are reduced by economies of scale: the more X you’re buying, the more efficient that purchase is likely to be. It costs less to offer a ridiculously nice physical work environment for your Nth employee than it does to offer that for your first ten. Based on what I know about negotiating power, etc., I’m willing to bet that those economies of scale remain true to some extent across other realms of company overhead and employee benefits.
Basically, cutting X% of a workforce will reduce a company’s costs. But it will probably reduce those costs by less than X%. Pay isn’t evenly spread around a company’s employees: the C-suite is far more generously compensated than other employees are. And trimming away the costs of benefits offered to that X% likely won’t reduce the total price of those benefits on a 1-to-1 basis, due to the aforementioned economies of scale.
So how much does it cost to fire someone?
In these layoffs, fired employees will receive severance pay. That’s a potentially squishy number, but according to Business Insider, Microsoft expects to pay $1.2 billion in severance for approximately 10,000 employees (averaging $120k per person). If you take Microsoft at their word about their severance package being generous—I’m reluctant to, but this’ll make my math easier—we could guess at a cost of $100k per fired employee.
We can also track how much work-time these companies are losing by firing these employees. What do I mean by that? I mean: there’s a time cost associated with bringing employees up to speed. It’s incurred at the start of any employee’s time at a company. To a lesser extent, it may be re-incurred with any shuffle of personnel.
Friends working in tech tell me it takes roughly six months for someone to be brought up to speed and usefully contribute to a highly specialized team, while ”a quarter or two” might be the normal range for other teams. At the six-month end, firing 6,000 people who’ve been at the company for six months or more is 3,000 wasted worker-years (for reference, Google has fired 12,000 people; Salesforce 7,000; Microsoft 10,000). Even if I assume everyone fired got up to speed in only one month (which isn’t borne out by the spread of teams that lost members), that still comes out to a cost of 500 worker-years for firing 6,000 people. That time was paid for already. Firing those employees, to me, looks like squandering that investment.
What is harder to quantify?
The above math assumes we only count the time of the employee being brought up to speed. It makes no accounting of the time and effort other people on their teams put into helping them catch up. I won’t try to account for that here. Any estimates there would be even more speculative than what I’ve done so far, and I think my assessment of the squandered time investment is already damning.
Another item I can’t quantify here is the cost of the institutional knowledge lost in this firing process. From others’ reporting, some of the people fired have been with their companies for up to 16 years. The company-specific and project-specific knowledge and expertise accrued over a decade-plus of work at a company can’t be understated; knowing who to speak to when seeking answers for specific problems, being able to tell others how older systems work, and having personal experience of previous solutions to prior crises all matter. As with the invested worker-time, all of that is being squandered.
Which brings me to another difficult-to-quantify cost which my friends in tech anticipate: decreased product resiliency. Because “automate yourself out of work” is the standard MO for many development teams, the repercussions of firing chunks of those development teams aren’t likely to be felt until a quarter or two from now when something inevitably breaks. When that happens, it’s a roll of the dice whether the person who designed the system and knew it best is still available within the company. The recent layoff wave, with its apparent scattershot approach, is well-designed to exacerbate those system failures by unpredictably removing the relevant expertise. Maintaining existing products and internal infrastructure will be harder, and that additional load will likely make future development more difficult as well.
All of this is exacerbated by the way in which the firings were carried out, which look a lot like intentional corporate self-harm from the outside.
Those laid off report learning of the firings by surprise. These people were frozen out of their ex-employers’ communication systems with no time given to pass on custody of any of their work to their coworkers. While many of these articles focus on Google, I’ve heard similar reports from people at Athenahealth. Any sane and competent organization—one that wants to preserve continuity of service, continuity of experience, and allow people to pass on custody of their projects to another person without leaving everything in a mysterious mess—would do this differently. Anyone picking up the projects of those who were fired will have to do their best to decrypt whatever notes made sense to the person who thought they’d return to their project the next day. It’s the opposite of good management.
Given that most investors look for companies that take advantage of market downturns to grow their business, this wave of firings seems short-sighted. It appears to have been conducted in a way that guarantees the erosion of product resiliency and handicaps meaningful transfer of projects from those fired to those who remain.
A final element I can’t quantify here: morale.
Companies, whether they like it or not, are communities. Communities function smoothly when their members are able to trust each other and predict each other’s actions. When a company fires 6% of its staff for no apparent reason, that damages any existing trust and puts the lie to predictability. It makes remaining experienced employees more likely to quit, or seek employment elsewhere.
Also, as a reminder, people may be friends with their coworkers. I am not in a position to quantify friendship. I can’t speak to what portion of the people fired had at least X friends, or whether they were real SOBs that everyone was glad to see leave (an ideal candidate for firing case #1). But seeing one’s friends lose their jobs for no apparent reason, seeing them suffer as a result, is disheartening, discouraging, and encourages anger and resentment—none of which are conducive to greater productivity.
Furthermore, I think this article about the pressure recessions exert on union organizing misses a critical point. Tech companies have been trying to quash union formation for years now. When tech companies show they are willing to fire people for no apparent reason, that may discourage and foster fear amongst their employees. But it can also be another incentive for workers to take actions their companies don’t approve of—like organizing. If they might be fired at random despite playing by the rules, what do they have to lose?
Scared people will certainly pay lip service to frightening authorities. But we have about two more years of Biden’s first term left to go. There’s no time in recent history when employees have been more likely to receive federal support in their unionization efforts. If these companies wanted to undercut pro-union sentiment, they chose a strange way to do so.
In summary, I don’t think these layoffs make much sense. The story tech executives are telling, of these layoffs being a sign of their seriousness and diligence and responsibility, simply doesn’t hold water. Even if dramatically reducing costs right now were necessary for these companies, the way in which these companies went through this firing process verges on self-harm.
My reasoning is as follows:
If a company over-hired, and there’s a recession coming, it makes sense to slow down hiring. It makes sense to reduce expenses. If the company were operating on thin margins and had little cash in reserve, that might require drastic measures and targeted cuts.
When a company is still making profit, and is sitting on a war chest of a hundred-billion-plus dollars, it can afford to eat into that reserve in order to conserve its existing expertise and expand its future capabilities.
Instead, Google spent $59 billion on buying back shares (see Ruth Porat’s segment of that Q4 2022 call). That increases stock price. It does nothing to improve a company’s fundamental performance.
Firing a swathe of employees largely at random has considerable costs. It incurs severance pay. It squanders prior investments of time and money. When firing at random, the money it saves in expenses does not match the proportional impact on institutional knowledge and capability. The firings damage worker morale and foster resentment. And the future cost of time lost to fixing broken systems without the people who knew them best, alongside the associated reputational costs when those systems’ failures impact the company’s customers, are injurious.
As far as I can tell, the only way that these firings make sense is if the companies involved expect a dramatic drop in revenue. In a moment of grim comedy, these firings may create the environment for those losses, and make weathering them more difficult.
But this brings me back to my suspicion that these firings weren’t about actual diligence and responsibility. The alternative is that these firings have little to do with the companies’ future performance, and everything to do with signaling to investors.
In response to lower than expected earnings over the last quarter, some investors are pushing for cost-cutting measures. This is fairly normal behavior. Investors who expect a recession are also pushing for companies to prepare, or at least to be ready for slower growth against economic headwinds.
How do companies show that they’re taking these concerns seriously?
By firing lots of staff, by creating a narrative of responsibility and seriousness, these companies are attempting to communicate that they’re sober, clear-headed, business-minded. As with the stock buybacks, this message is a reassurance that the company cares about their investors and will do what it must to improve their stock value. But this privileges short-term behavior over long-term investment.
As someone who owns some shares in some of these companies, this is disheartening. These are bad decisions. They’re bad decisions that have been executed poorly in the most injurious ways. They look to me like a surgeon grinning, hands smeared with blood and full of mostly healthy freshly-excised tissue, reassuring the audience that the cuts they just made were all well thought out. I can only hope the audience is not convinced, and doesn’t reward this behavior.
I have few doubts that these companies will weather our next economic storms mostly intact. They have many billions of dollars in reserve, they can afford to choose poorly again and again. But this? This was a mess.
It’s also an opportunity. Those workers still employed by these companies can make their voices heard. Even if they still believe that the company they work for earnestly shares their best interests, they can organize and push their companies to focus on long-term planning instead of the market-rewarded sugar-high of quick fixes and flashy cuts.