There's been a lot of discussion of this in IT circles. Here are two articles discussing effects of this by two people who are in the thick of things:
http://startupboy.com/2011/12/13/why-you-cant-hire http://www.paulgraham.com/invtrend.html
For a number of reasons, the capital costs to start an IT company have collapsed. Back in 1996, pushing 2 terabytes of data to the Internet a month at 10 MB/s cost $50,000 a month at a discount ISP. Today I can do that for $20 a month, and get a virtual server thrown in for free. The cloud/VPS companies like Amazon and Linode are in a vicious price war, so even that price is collapsing while the speed/bandwidth/memory etc. increases.
Also, in 1996 if I wanted to write software for people's personal computers, I'd have to work out a deal with a software publisher like Brøderbund, who would eventually package the application and ship it to stores like Egghead Software, where people would buy it. Nowadays I send Google $25, wait for them to approve my account, and within minutes of that can publish a program for the use of the 1 billion Android phones and tablets outstanding. Same thing with Apple.
It's all anyone talks about at startup tech Reddit knockoffs like Hacker News. It all happened fairly recently also. Even Facebook was saddled with costs for servers and bandwidth in its early days that would be considered relatively enormous to what they could do it for today.
Then throw into the mix all of the open source software out there, accumulated on places like Github.
Look at Instagram. Started in October 2010 with $500,000. 3 months later they raise $7 million. Within 17 months of starting they had 100 million people using it and were bought out by Facebook for $1 billion. They had 14 employees when they were bought out.
There is just a mass of constant capital spent costs existing in this space from established businesses. For new businesses, constant capital costs are next to nothing, almost everything goes to variable capital costs.
What's happening is every semi-talented engineer with a little business savvy has started his (his) own startup, perhaps with another engineer friend or two. Maybe a designer. The market is flooded with angel money (investment money for very small businesses).
With IT, quantity is transforming quality, and the effects of changes in production that have been happening for years are finally bearing fruit. This is throwing relations of production into an upheaval.
The question is, why would a decent engineer nowadays want to work for a company that he did not own at least 15% of, if not more? We have direct access to a consumer market of billions now, for almost no cost. As I said, even the costs Facebook had to pay in 2004 for servers and bandwidth, which they had to raise angel and VC money for. Things had changed enough that the VC's did not have enough control in 2006 to force the founders to sell the company for $1 billion. And capital's power, insofar as founder/VC struggles, have only diminished since then.
The market is flooded with low-level angel investment (< $1 million). There is something of a capital strike though ("Series A crunch"). Though companies with incredible "traction" are funded by VCs. The capital strike isn't having much of an effect though, engineers are realizing that at the moment they don't need capital investment. The San Francisco real estate market is booming, so I hear. But firms aren't really getting VC, never mind IPO's. Really, tech companies don't want to IPO any more. It's why Facebook didn't IPO until it was worth $104 billion. They see no reason to leave private markets.
But getting back to it again, there is a tech capital strike, which normally kills economic activity in the market. There is a capital strike because the rate of profit is being threatened. VCs are trying to force millions of extra dollars onto hot companies because they still want a 20% share of a company on a series A round. But hot tech companies don't need that capital nowadays. Unless they're building artificially intelligent self-driving cars like Google is. Normally a capital strike forces terms back onto big capital's side, but due to the existence of a massive constant capital infrastructure from established companies, engineers and new companies don't really need capital, considering the angel small capital markets are flooded with money, and they can just set growth to go with revenues and profits after that, since pretty much no one is getting funded in the millions. And if the company fails, start another one, there is so much angel money floating around out there.