Memo to Tim Cook: not sure you'll want to have $100+ billion dollars just chilling in the bank when ex-employees start winning class action suits against you.

Suppose developers/software companies create a large consumer surplus which they fail to capture. If prices go up, then either quantity supplied will go down (reducing consumer surplus) or companies will focus on efforts where they can capture a larger portion of the gains due to trade.

In practical terms, low cost developers might make a free music service economical. If the cost of producing software goes up, the free music service might be uneconomical, and the developers may refocus their efforts on creating DRM (allowing producers to capture a greater portion of the gains due to trade).

I'm not asserting that this will happen, merely pointing out that it's not hard to imagine scenarios under which rising developer costs might be bad.

[edit] Allow me to expand on that in a very specific way. I'm currently sitting in Bengaluru, surrounded by some excellent developers. What effect does wage inflation in Silicon Valley have on people here? Every time a Silicon Valley engineer's salary gets inflated by US$20K, how many people here are affected by reduced job prospects or raises? From that perspective, what's hard to see is how a bidding war among SV employers could possibly be a good thing.

http://royal.pingdom.com/2009/05/14/congratulations-google-s...

Also, dividing revenue per employee by salary for a subset of employees doesn't really give you the right ratio. A lot of corporate revenue is not directly tied to developer contributions, and it's a rare employee who actually moves the bottom line by 10x their salary. At best you could divide the revenue attributable to that year's development by the salary paid during that time, but even that's a weak approximation.

Wouldn't it be a neat if user profiles here displayed a downvote/reply metric? Not replies in general, but replies to comments which the user had downvoted.

You start out claiming that developer salaries in SV are out of proportion because SV developers do not create enough value. You also state that if developers are paid too much then there won't be enough left over for other workers, non-technical and such. Of course, for that to be true we'd have to assume that hiring an engineer in Silicon Valley is a net loss or a wash at best, because otherwise hiring one actually means more resources for further hiring and salary increases.

Then, when confronted with the fact that they most certainly do create proportional value and at a rather healthy ratio no less, you instead make the claim that they are overpaid because cheaper work can be had elsewhere (in Bengaluru). You make the implicit assumption that, for a firm based in Silicon Valley, an engineer in Silicon Valley is equivalent to an engineer in Bengaluru, India (or anywhere), provided they are equally talented. Not many of your readers will grant you that point. At any rate you seem to claim the market should determine what the fair salary is, when the very issue at hand is collusion by several large employers in Silicon Valley to distort the job market there. Is there any force acting in the opposite direction that you think is raising salaries there? Did all the engineers there unionize when nobody was looking? Because otherwise developer salaries are determined by market forces. Everywhere. (Again, with the issue at hand being an exception, but an exception not in favor of your argument.)

Your principal beef seems to be that developers in India don't make enough money compared to their peers in the States and specifically in the Valley. You make the claim that rising salaries in Silicon Valley mean lower salaries and reduced job prospects in Bengaluru. You do not justify this claim nor do you give anyone any reason to give a shit in the first place. Let me make an equivalent claim: rising salaries in the Valley mean lower salaries for everyone everywhere else. See now why it's bullshit?

I was certainly never confronted with the "fact" that SV engineers "certainly" do create proportional value. I was confronted with a bunch of bad math that proves nothing. Did you miss the part about the target ratio not needing to be 1:1? Or the part about developers' contribution during a year not being the only (or even main) determinant of revenue during that same year, which makes division of revenue for all by salary for some utterly ridiculous? Where are these engineers who drive the bottom line by 10x their burdened cost (not just salary BTW)? I've been one myself, I've met others, but for every such person I've also met dozens who contributed 10% or less over their cost. You need to stop assuming, and start measuring.

OK, on to the point about whether SV engineers are worth more than those elsewhere. Let's forget Bengaluru for a moment. Consider Boston or Boulder. Brisbane or Birmingham. Engineers there are every bit as good as those in SV. Sure, there are costs involved in having developers who are elsewhere, but that's balanced by the extra cost of building out offices etc. in one of the most expensive markets in the world. Buy some decent teleconferencing gear, provide a generous travel budget, and you'll still come out ahead even in these relatively high-salary locations. Now consider a place like Bengaluru. I haven't worked with everyone here, nor have I worked with everyone in SV, but I've worked with enough to be skeptical of the assumption that the average SV engineer can deliver enough to be worth several times the salary. Some can, most can't.

That's two strikes against you. Let's go for three. Should the free market determine wages? Absolutely. Do these no-poaching agreements subvert the free market? Again, absolutely. Does that add up to a conclusion that a bidding war would be a good thing? Absolutely not. CDNF. A temperate amount of poaching would be a good thing, but I don't think that's likely to happen. What's more likely to happen comes back to your statement about "a net loss or wash at best" because at some point that does become true. I personally think it has already become true at many companies. Those newly hired engineers are going to count against this year's budget; their contribution to revenue won't show up until next year's. The extra cost of hiring that first developer of the year has to be made up somehow. Part of it will be made up by reducing capital costs and travel budgets and such. Part of it will be made up by hiring nine new developers instead of ten (wherever they are) or by giving smaller raises to current employees (including developers). Maybe that "rock star" you just hired away from Google will pay off next year, but this year (and probably the next couple) their bidding-war premium will exert downward pressure on everyone else. A bidding war is not only bad for developers elsewhere, or for non-developers. It's bad for other developers in SV besides the few who win the lottery.

The problem with the no-poaching agreements is that they've created pent-up demand. Now, like soda that has been shaken, the newly opened bottle is going to fizz all over the place. Those making the agreements should have been prosecuted years ago, because waiting has created an even bigger mess. The absolute least likely outcome is an efficient short-term reallocation of resources. More likely is that some companies without deep pockets will be starved out of the market.

When I say your position is all over the map, I mean you can't seem to decide what is fair for you when determining what a person should be paid. On the one hand you're talking about the proportion of value to salary and target ratios, and on the other you're saying that the market should determine wages (as opposed to does). The fact is, if you find yourself in a position where you need to get something done, and the only people who can do the job are demanding a salary only 10% lower than what you expect the project to make, then all that value ratio bullshit goes out the window, because 10% is better than 0. If you've got to lay people off, then you can look at who is contributing what. Maybe you can do that when hiring too but it's much harder. But we're talking about the labor market in general, not in individual cases. The fair wage, that is to say the right wage, is the one you can get in a free market, and the topic at hand happens to be collusion against that market in a direction that is not favorable to your position. Again I ask, what is making them high?

At any rate, at Apple, Google, Facebook, etc., the margin is a hell of a lot higher than 10%. All this talk about whether someone in Boston or Brisbane or Bengaluru has a skillset comparable to a more highly-paid developer in SV in pointless, because they aren't in SV, and SV is clearly where these companies want them. If you think they're still paying too much, then you should find a way to short the market for software engineers in Silicon Valley. Really.

Anyway, replying to you certainly didn't take any balls at all, it just took time. I happen to have time in abundance for the moment so I used some of it to try to make a point. And mostly I did that to check that I still could, as a bit of a challenge. The challenge stemmed not from the fact that you put forward a good argument, but rather that your argument is so poorly constructed that it's hard to form a really coherent response to it. I am now also intrigued by your persecution complex which is why you're getting this second reply. Look, there isn't some downvote mafia you've pissed off, and the people who downvoted you aren't cowards. You've put forward a provocative thesis, for this place at least, but you don't get downvoted for that here. You do get downvoted for putting forward arguments that lack substance, especially so if you have an attitude about it.

WAT?

Top software developers are very well-paid. No complaints on that one.

That said, the value-add of a good software engineer (not a "rock star", just 95th-percentile, which is average in some communities, e.g. functional programming) is easily half a million a year.