I'll add an executive-level twist based on labor cost strategy to the general ideas about demand and supply at the engineering level. For emerging markets, this may be more unpredictable and a bigger factor driving decisions than pure talent competition. Talent markets do not operate in isolation of other business concerns.
There may be unmet demand, but it is demand to fuel something that has been judged to be a "Ponzi scheme until proven otherwise." A potential bubble economy in short.
Consider the environment in which a CEO of one of the larger companies has to make talent management decisions at a strategic level.
Facebook's business model is unclear. Microsoft is struggling for relevance in a clouds-and-tablets world. Google's search core is aging, with spam pollution, Android still makes less money than iOS for Google, Amazon may be running into middle-age cash management problems soon if rumors are to believed, and given its aggressive loss-leader approach to developing new markets like the Kindle.
Even Apple (a relatively "old economy" company due to its hardware-centric business) is struggling with online economics and strategy (mapgate etc.). Its devices may still be great, but it is has never been clear whether Apple is good at online economics, or whether its devices are propping up bad online thinking.
In this environment, nobody is entirely sure whether this whole sector is a minor one, a hyper-inflated bubble, or something substantial that will account for 50% of GDP or something in 10 years. If you are the CEO or high-level exec at one of the large companies driving the game, you cannot just think in terms of how much you want to pay for talent.
You have to decide how much you want to pay for talent, adjusted for the risks of the market opportunity you are going after.
Why? Because labor is the biggest cost fraction here. Take a simple case of a company with 50% s/w labor costs, $10 million total costs, and $20 million revenues. So 50% gross margins, and this company spends $5 million on labor now.
The board and CEO think they could hit an addressable market of $100 million (5x) if they scaled their engineering correspondingly (5x). If they scaled proportionately (something biology teaches us is never the case, but never mind that), they'd be spending $25 million on labor.
Let's say you're willing to take a $100 million market at 25% gross margins, which means if your cost ratio (labor:everything else) remains the same, you could actually afford to budget an additional $12.5 million for getting the best talent. If you spread that evenly across old and new hires, and assume for a moment that everybody gets paid the same, you suddenly have a way of offering a 50% premium over the competition (I think I did my math right here...)But if the market opportunity comes with only a 50% chance of being realized, with a 50% chance that the market cannot be grown at all, where does that leave you?
You grow aggressively on labor, and if the plans don't work out, you have to do layoffs and get yourself into deep trouble. Your reputation may take such a deep hit that even your original base that you were growing from is threatened, because the talent you do
have loses morale and starts quitting in droves for the "next cool company."
What's more, you have to take into account that talent prices may drop as supply from the education system slowly starts to catch up. If you have a 5 year plan to grow 5x, and are betting on an exponential profile with labor scaling towards the end, you want to dollar-cost-average your talent acquisition accordingly, and get the cheaper/more plentiful new grads to the extent possible.
Talent management is equal parts microeconomics, culture management, business model risk management and plain old supply-chain smarts. A startup can afford to play only one of those games, but if you run a company of any reasonable size (like 300+) you get yourself into trouble if you are too simplistic about it.
I forget which question this was, but somebody had a brilliant answer about the pathologies of a company that raised too much money and then hired too many people who sat around twiddling their thumbs and partying essentially.