Hacker News Reader

Comment by Apocryphon | Hacker News Reader

A tech union doesn't necessarily have to behave like traditional unions that favor seniority, and not all unions prioritize seniority anyway.

View on HN

It doesn't have to, no, but it probably would. Certainly most unions seem to.

[EDIT] I'm poking around the sites for various unions for technical workers and engineers outside the software industry and, at a casual glance, many seem to factor in seniority as well. Are there examples of technical or engineering unions that are not seniority-based? Seems like those would be better indicators of what a software workers union would look like than the usually cited actors or sports player unions.

Why "probably"? A union is whatever its members make of it, and most engineers at any given tech company have relatively little seniority. It's hard to imagine them voting to join a union that would so obviously not prioritize their interests like that.

My guess is in most cases you'd still end up with an avenue for merit based layoffs, they just wouldn't start until new hiring had been completely stopped, and perhaps include an option for everyone to take a pay cut instead (which might be bad! but it's an idea I've heard floated by engineers way more than seniority based protections).

What's wrong with employees just voting with their feet, instead of mandating rules like this? If a company behaves horribly just go somewhere else

Doesn't seem to happen in tech judging by Amazon horror stories about PIP culture and the like.

I think it's pretty easy for an average US citizen or green card holder engineer who gets pipped at Amazon to get hired elsewhere.

The people who really get screwed are the OPTs/H1-Bs who have a very short grace period to get a new job or leave the country.

Sure, but 1) it goes to show that tech workers won't necessarily "vote with their feet", 2) in the event of an industry-wide downturn, that ability to vote becomes far more limited.

Amazon pays software engineers very well (compared to software engineering jobs in general). Perhaps people don't vote with their feet because they like the money.

Amazon comp was entirely predicated on stock, wasn't it? Because they uniquely have a salary cap that's not particularly high.

https://www.levels.fyi/blog/amazon-salary-negotiation.html

It seems like there is more Amazon non-warehouse employees interest in unionizing now anyway:

https://www.businessinsider.com/amazon-layoffs-corporate-emp...

Not exactly - while the initial stock they grant vests on a 5/15/40/40 schedule, they also have a large cash "bonus" which "vests" monthly for the first two years. With their assumption of 15% stock growth per year, you should have equal TC per year for your first 4 years. Of course, assuming 15% annual stock growth is silly, but you can run the numbers however you want. The point is that compensation for the first two years is predominantly cash.

No? I've definitely heard stories that Amazon is having a harder time hiring people because many workers are churning out.

> an option for everyone to take a pay cut instead (which might be bad! but it's an idea I've heard floated by engineers way more than seniority based protections).

We already do that, it's called RSUs. If a portion of your compensation is in RSUs and the market goes down, you automatically get a paycut. Better than being laid off, isn't it?

This is like doing the same thing but expecting a different result. You think unions follow some optimal logic but I’ve seen no examples that do. The only real expectation you can have is that it would end up roughly like existing unions.

I have had personal experience with men in leather jackets giving you the “better vote x or you might have an accident” unions. So I’m a bit biased.

> This is like doing the same thing but expecting a different result.

Tech culture, and Silicon Valley in particular, is so obsessed with recreating things from first principles and experimenting for the sake of it, that you would expect that they'd give reinventing unions but under a different name a shot.

There are a few problems with unions, many of which can be summed up as principal/agent problems. But another issue is that unions are fundamentally democratic, which leads to some misaligned incentives.

Seniority is a completely legible and unambiguous criteria, and it’s one that most workers can eventually benefit from in the long run. Merit is more of a judgment call. Management at least has the incentive to get that judgment call right (which isn’t to say they do get it right, but at least they have the incentive to, modulo the same principal/agent problems that unions also introduce), but unions don’t. So if I’m part of the top 20% of workers at a firm, or part of the top 20% that the firm could reasonably hire, my incentives are actually better aligned with management than with the other 80% of workers who would dominate the union. The classic union solution to this problem is closed shops, but that’s hard to achieve without specific legal and regulatory moats.

Also:

> My guess is in most cases you'd still end up with an avenue for merit based layoffs, they just wouldn't start until new hiring had been completely stopped

This is another common failure mode of unions. People who already work in a given field are in the union, but people trying to break into the field are not. So it becomes harder to hire new people, which actually leads to longer tenures and makes seniority seem like a better deal for the union members.

I tend to imagine that tech unions would be different from organized labor cultures in the past. Partly because it's a newer industry with different culture, and actively seeks to use technology to try to invent new social relationships, from the FOSS movement to the gig economy. But perhaps that's being overly optimistic towards tech/Silicon Valley's pretense of disruption, innovation, and doing wildly different things from the past, and naive towards its hypocrisies.

What way would be ideal? It would have to be an objective measure applied universally. I challenge you to come up with a reasonable measure. In reality, front line workers and managers (the good ones) know ability of people and how much they contribute. In my experience it was very obvious and agreed upon for the most part. But none of the (in)effective people could be discerned in an objective way.

My point isn't that unions could necessarily protect layoffs, and thus I am agnostic about how they should respond to layoff criteria. My point is that unions could protect benefits that might go away in the event of a market downturn, when tech labor power is similarly weakened.

Right, but everything has tradeoffs. If the org chooses to protect benefits, then it may mean more cuts. For instance, to save $X a non-unionized firm may have to cut 5% of the workforce, and it would likely choose weaker performers. But if they have to provide more benefits and severance to those they let go, they may have to cut 10% of the workforce instead. In my experience layoffs were always measured. The people affected were usually weaker performers or in functions that weren't practical anymore (e.g. recruiter during a hiring freeze).

For me the tradeoff is not worth it. I understand that there will be ups and downs and I may one day be a victim. But I appreciate the flexibility that likely led me to be hired in the first place without a traditional background in tech. Basically they took a chance and luckily it worked out. But if the firm had a much higher cost to get rid of someone, they likely would have went with a more traditional candidate

By benefits I don't mean simply TC pay package items or employee perks, I mean not being subject to games industry-style unpaid overtime crunch, or forced RTO for those who had moved hours away from the nearest office during the pandemic, basic QoL "luxuries" who might be put into jeopardy as labor power weakens in poor economic situations.

Reply on news.ycombinator.com