Now that users can do almost anything (simple) on a computer or even their phone, they now expect that anything they can imagine (vaguely, inarticulately, even impossibly) should be easy to do.

Unless you're at one of the rare shops that's well funded and not directly dealing with users, you will likely be in a no-win position.

Deliver a flawless system and you go unnoticed. Instead, you get asked "can it do this ?"

Or worse and most likely, you step into a position with an existing product that you have to continue development of. It will be behind schedule, over budget, and a complete architectural disaster. What's more, it won't match what the users need because nobody bothered to dig deeply to find out what the users really needed (as opposed to what they initially said they wanted - there's a huge difference).

Am I bitter, yes. I'd rather be a lawyer. At least then I'd still be getting rich doing crap work.

What you describe does not surprise me, and in fact much of the blame does go all the way up the US management chain.

Ultimately the US business game is about quarterly earnings per share, and using stupid tactics like these help boost those numbers temporarily. Of course in the long run overall performance of the company suffers (and along the way, Americans have fewer job opportunities).

My personal experience with outsourcing people is that they simply cannot think outside the box. In fact, they cannot think beyond explicit instructions. For the level of energy it takes to adequately instruct an offshore worker, we can do the work ourselves. This is especially true with the more agile languages and toolkits these days.