It's unclear whether that would be better. In control theory, underdamped systems are generally preferred as it allows you to reach the desired value rapidly even though you often overshoot.
Overdamped systems (IE hire in moderation) can be very slow to respond and often spend the majority of their time in the un-desired regime.
It's quite possible that tech over-hiring when people were forced to work from home yielded a net good for society even though many of those hires were let go a couple years later.
He's advocating that the powers-that-be should consider targeting a slightly underdamped hiring system. Where they over-hire, and as they're overshooting the goal then reduce the rate of hiring, allowing the normal low rate of attrition to bring the PV back down to the SP.
e.g. a one-half critically-damped system. something following a standard first-order ODE like:
where h is the instantaneous number of employees, h* is the desired number of employees, f(h) is your attrition rate as a function of current number of employees, and Kp and Ki are "rate of hiring" coefficients which will be your proportional and integral terms of a PID controller to control your rate of hiring. The point is, the "rate of layoffs" here is zero and doesn't appear as a term.
Is this a proper model for a human organization? Who the fuck knows but it's the model you're advocating for and I'm just re-explaining what the other guy you replied to was suggesting in the first place, but this time using your framing. Yes, he also mentioned they should consider a slightly over damped system but he 100% did advocate for considering an underdamped system.
Ehhhh, your argument was more that "severely underdamped" could be the best, even if it requires an additional forcing factor (firing/layoffs) to bring it back down to the setpoint.
The person you replied to was arguing that perhaps something a bit more mildly underdamped could get rid of the requirement for that extra complication (firing / layoffs).
Yeah, and I disagree. Goal should be to get to the setpoint as quickly as possible even if you overshoot as human nature is to keep the status-quo which generally means talented people wasting their time at obsolete companies.
Control theory in sociology proposes no such thing. Your using an engineering term used to describe machines which is largely the point; human capital being as disposable as a cog in a machine.
Tech companies over-hire to excess and then under-hire to excess. It's like people trying to follow a 1000-calories/day diet who then give up and eat 4000 calories a day. They are never going to hit a healthy weight. Neither is my jackass company, who has laid off more people in 2023 than we hired in 2022, and is still on a hiring freeze.
Honestly, I think this is a wicked problem, in that it's impossible for us to know in advance what the right decision would have been. Was it foreseeable that the economy would be volatile post-pandemic? Yes. Was it guaranteed that it would mean layoffs? No.
In addition, what about the boom itself? What if you're cautious but, during that boom period, you get outcompeted in key areas as a result? This might put you in a worse situation during the bust phase than if you'd over-hired.
Hindsight is 20/20, so I prefer to avoid being judgemental over these types of decisions, while still holding people accountable to doing the right thing, even in tough times.
Or lots of individuals just don't get hired and you'd have had lots of people railing against headlines like "Despite surging demand, companies keep hiring choked off."
To use an even more reasonable framing, absent some small numbers effect or some really high-impact shock, it's incredibly stupid to almost double the size of a company in one year.
That's how you lose your culture and throw the dice again on market fitness. (Well, that and randomly firing people.) Knowledge work just can't scale that fast, and nowadays every work is knowledge work.
What those companies did is bad from every possible point of view.
It's unclear whether that would be better. In control theory, underdamped systems are generally preferred as it allows you to reach the desired value rapidly even though you often overshoot.
Overdamped systems (IE hire in moderation) can be very slow to respond and often spend the majority of their time in the un-desired regime.
It's quite possible that tech over-hiring when people were forced to work from home yielded a net good for society even though many of those hires were let go a couple years later.