Comment by hardwaresofton | Hacker News Reader

Comment by hardwaresofton | Hacker News Reader

Parent Full thread

> why do you only compare to Japan and EU? American engineers get paid more than anyone else (except for EU outliers). I make slightly above 100k in Australia, it's a bit tight with the housing situation and cost of living but it's liveable. Cost of living in the US is lower except for maybe healthcare but engineers get companies pay for them anyway.

Well those are just the biggest/most relevant tech labor markets I could think of, roughly by GDP. I don't know much about Australia but couldn't comment there, thanks for the insight.

> I don't have insights into other engineering, but I think software engineers are overpaid. Would I enjoy being overpaid? Of course, but it's hard to say it's fair pay considering what people in other professions are making.

I don't think comparing pay across professions makes sense -- why compare a plumber or doctor to what a software engineer does? the outcomes are completely different, the risks are completely different, the upside/profit potential is completely different.

I lean towards considering software engineers as underpaid because the power of automation (whether software or hardware -- because you generally need some kind of soft/firmware to make robots move) is insane. If software engineers were ever paid a percentage of the revenue impact they generate (which is not how markets work, but I digress), software engineers would be paid much much more. I think it might be possible to compare across professions if you think of % of value creation captured let's say but it gets tenuous.

In the end the only "fair" determination is what supply and demand based markets dictate. They're not perfect, but they approximate other forces/influences decently enough. When those aren't enough, governments may need to step in to enact policy.

fair enough, fyi my line of thought is a bit Marxist, you should get paid based on your labour, time, mental energy etc put in the work as much as the result. When you pay people according to only "impacts" you create an incentive to make "viral" things. In software these days even without such incentive taking full effect, over abstraction, one-size-fits-all, github star counting are becoming bigger problems.

> you generally need some kind of soft/firmware to make robots move) is insane

The fact that you completely ignore the hardware is very telling. CPU has been improved according to Moore's law in the past few decades, if anything modern software engineering failed to live up to that potential and has been free riding on top of progress in hardware.

It's insane to make robots move but software is a small part of that, there is also mechanical/electrical engineering, battery techs etc. Software sitting at the top of that stack shouldn't takes all the credit. It's very hard to innovate on the lower layers with all the lab work and physical constrains.

> fair enough, fyi my line of thought is a bit Marxist, you should get paid based on your labour, time, mental energy etc put in the work as much as the result. When you pay people according to only "impacts" you create an incentive to make "viral" things. In software these days even without such incentive taking full effect, over abstraction, one-size-fits-all, github star counting are becoming bigger problems.

Hmnn but doing it that way feels like it would devalue doing things smarter which seems to be more beneficial to society. Single-mindedly incentivizing labor seems like it would skewer efficiency -- I guess like most things the right answer is in the middle.

I'm not sure what you mean by "viral" things but I'm trying to discuss only about work output. If the work output is virality (let's say a social media manager), then it makes sense to measure that way, but usually businesses don't deal in that currency -- they deal in product shipped out or contracts landed, etc. That said, I'm well aware that systems that think themselves to be closest to the capitalist ideal are pretty brutal on the cogs in the machine.

As a side note, it's a red flag for me when I see github projects that are too concerned with stars/going viral, or even when I see presentations about projects that mention stars. Everyone thinks it's silly but everyone keeps doing it -- if only we'd stop then people would start judging projects on things other than adoption. Then again, adoption is the biggest metric for most people, because it generally is a proxy for "availability of free support/feature development", and in that sense I'm guilty of the same.

> The fact that you completely ignore the hardware is very telling. CPU has been improved according to Moore's law in the past few decades, if anything modern software engineering failed to live up to that potential and has been free riding on top of progress in hardware.

> It's insane to make robots move but software is a small part of that, there is also mechanical/electrical engineering, battery techs etc. Software sitting at the top of that stack shouldn't takes all the credit. It's very hard to innovate on the lower layers with all the lab work and physical constrains.

Agreed! There is more than enough value to go around to be shared between all those kinds of engineering -- just because I didn't mention lithography engineers or materials science engineers or people who smelt silicon doesn't mean they don't exist and don't deserve to be paid.

Viral or scalable in business/tech terms, means things that can be sold multiple times. Often times the value is already created and rewarded. In our economy we're not only incentivizing value creation but also building walls around innovations and gatekeeping the fruits of them, so that the innovators, the software developers like you wish, or whoever can earn royalty while no longer innovating.

Yeah, I get where you're coming from but I can't really agree that incentivizing raw labor is the way to get there (to a more equitable outcome for labor). The capital/labor struggle is a deep deep topic, but I think we can say at this point that with automation becoming increasingly common, labor is losing it's leg to stand on. We're going to get a lot farther by reining in capitalism than trying to return to piece work payment.

> I can't really agree that incentivizing raw labor

I read history books, I'm not that stupid.

The reason I bring this up is a couple of comments up you said software creates tremendous value, I guess this is true for many other types of innovations too. You proceeded to say software developers need to be paid proportionate to that value and I disagree there.

If we look at salary as incentive I would say that software engineers get enough incentive to do our job. As long as we get paid better than most other professions, society/the economy don't need to pay us more.

While pay should not be based on labour, it should be capped, probably based on the labour. Otherwise every cog in the machine earns some royalty (infinitely, if the machine scales infinitely) and there is no money left for "non-machine".

The bad thing in this industry is work/life balance, we just need to fix that. Although the pay in some countries are atrocious, I'd say that software engineers in most western countries are fine.

> I read history books, I'm not that stupid.

Didn't imply you were! I am assuming the best as people do here on HN.

> The reason I bring this up is a couple of comments up you said software creates tremendous value, I guess this is true for many other types of innovations too. You proceeded to say software developers need to be paid proportionate to that value and I disagree there.

> If we look at salary as incentive I would say that software engineers get enough incentive to do our job. As long as we get paid better than most other professions, society/the economy don't need to pay us more.

> While pay should not be based on labour, it should be capped, probably based on the labour. Otherwise every cog in the machine earns some royalty (infinitely, if the machine scales infinitely) and there is no money left for "non-machine".

So I still disagree with this point, but I want to point out that I am in favor of the idea that every cog in the machine earns a royalty on every piece of the process. I believe both machine and non-machine could win in that situation. If you have a new idea (and the funding to go after it, etc) then you get to decide the cap table essentially -- so it's up to you to set rates on the royalties and up to labor to agree/deny. How you'd work out royalties in relation to work done/etc is difficult, but it's something that feels like it can be worked out. Pensions used to be a thing, and they're kind of like that.

> The bad thing in this industry is work/life balance, we just need to fix that. Although the pay in some countries are atrocious, I'd say that software engineers in most western countries are fine.

Weird, I think that work/life balance for software is actually really good -- you're not going to find much work/life balance (or companies that even pretend to care about that) in the blue collar job market. The ability to up and move with transferable skills is really remarkable in tech as well... In short I agree with you that tech has it good, but I don't think it makes sense to restrict that "good" based on the "bad" of other fields -- we don't need to be equal. If workers are getting exploited outside of tech, let's make them less exploited and get them what they're worth -- if tech workers are exploited let's get them what they're worth.

The question of how to determine that worth is the big difference we're at loggerheads about I think, and I don't think it makes sense to judge by amount of "work" done. That very idea is rooted in some sort of moralistic argument that work/expenditure is good/virtuous/worth being paid for, and I don't think that's a shared value. Labor is given meaning by the fruit it produces, and that's how I'm trying to judge the value of labor -- the value it produces (whether immediately or down the line) -- royalties do that rather well compared to our current system.

> software creates tremendous value

According to what metric? I suspect most of the stellar SV salaries come from the economic model.

> If we look at salary as incentive I would say that software engineers get enough incentive to do our job. As long as we get paid better than most other professions, society/the economy don't need to pay us more.

I disagree. The value created is going to be accreted somewhere, and I would prefer that somewhere be the labor that created it rather than the capital funded the creation.

Should you be accredited for the Internet, PC and other things, would that be fair? I think it's a delusion. You write some code that do a specific thing. It's the whole infrastructure and economic structure combined with that piece of code and creates such a massive value, it can't be acreditted to just you.

I think what is happening is that people in general, not only software engineers, are getting paid much more and sometimes less because of increasing impact due to tech/globalization scaling and winner takes all network effects.

Look at how much some YouTubers make. Due to the Internet, they can reach far more people than before and make far more money. Also, their popularity begat more popularity without commensurate additional work on their part.

The same is happening to tech companies. You have that same power law with a few at the top making a lot of the money. It applies to YouTubers, tech companies, software engineers at those tech companies, etc... The rise of tech billionaires is another manifestation of this.

Of course this is creating a lot of pushback from other parts of society that are sometimes actually losing influence, money, etc... since there is a negative feedback loop on the other side.

> fyi my line of thought is a bit Marxist, you should get paid based on your labour, time, mental energy etc put in the work as much as the result.

You can work very hard (dig useless ditches with shovels) or on very tricky things (write about things nobody cares about, nobody understands and do not matter) and not make things of much value. A society that does that in aggregate is a society that either dies or falls behind.

Reply on news.ycombinator.com