Yeah 2002-2005 or so the conventional wisdom is that there wouldn't be software written in America anymore -- that companies were going to beat a path to Asia just like they'd done in textiles and furniture.

Economists explained that software is a highly "tradeable good" and Asia -- especially India -- was filled with millions of highly educated cheap programmers.

Given that the number of people employed in software in America has remained constant since 2000, and that wages are about where they were then too, I think the evidence is that there has been a substantial "pressure release valve" effect going on. So while the most dire predictions of the 2000s haven't come true, it's also true that the confident optimism of the 1990s ill advised.

I would say that software development has been good to me so far as a career, but I still think I would hesitate to encourage a young person in America to pursue it.

> I would say that software development has been good to me so far as a career, but I still think I would hesitate to encourage a young person in America to pursue it.

This is absolutely terrible advice. Please do not ruin some young American's life with this advice. Technology seems to be one are where the demand seems to always outstrip supply, and a person trained in Computer Science would be capable of learning new skills and advancing rapidly in the tech industry.

I think that software is a unique industry in that it is a "self growing" industry. Unlike other industries, in which the more workers there are, the lower the pay, I think with software developers the opposite is true. I believe that the more software developers there are, the more things will be automated, and therefore the demand for software will grow with the supply.

Furthermore, since it the complexity of software grows exponentially rather than linearly, I believe that as the number of software developers grows linearly, the demand will grow exponentially ;).

San Francisco seems to be the only place where that holds true, and I say that not having tried to get a job there. The average length of time for engineers I know of trying to get a new job is 3-4 months. Any one of us could easily be a shit engineer but its in the 10s of people now. While that's better than a lot of industries now, that doesn't mean that demand outstripped supply. Its just at an ok point. If demand was much higher you'd see salaries rising rapidly at least and I know of few companies that are willing to give significant raises to keep someone vs hiring someone new. Why else is there advice to job hop so frequently to increase your salary?

Giving raises to existing employees is not related to demand and supply at all, but to other institutional reasons. Managers have more freedom to give offer higher salaries to attract good talent, than they have to offer super high raises, which can also cause discord among teams unless you offer raises to everyone, which may be more money than the manager can afford.

Of course it depends on the market, but I disagree that is just SF. Seattle, NYC, SoCal (LA, San Diego etc.), Boston, Austin all have a thriving tech sector, although Bay Area still far outstrips them in demand and compensation.

Those other factors due exist for preventing raises to other employees. The claim, however, was that demand was _far_ exceeding supply. When that happens with any good people start hoarding the resource and going out of their way to maintain their existing supply. If there was truly that great a difference between the demand for software devs and the supply of them, we'd see companies doing their best to keep their employees, at least on average. Instead we're in a situation. Where companies are engaging in some sort of brinkmanship with their employees to see if they'll accept the lack of compensation increase, and they dont appear to care if they are loosing employees due to it.

I completely agree with atarian.

I'm a bit of an oddity in that I took summer / nighttime college courses in 1995, when I was 10, then had a job programming right during the dot com bust. Everyone said that even if tech recovered it would probably all go to India so when I entered full-time university I went so far as to study structural engineering, something I didn't think would be easy to outsource.

When I transitioned back to software I was actually shocked at how big of a skill difference the median Indian and American / Canadian actually had. The top 1% between the two cultures is similar, but the median Indian, circa 2009 when I last interacted with them, was far behind.

Maybe academia was painting a different picture than the corporate side. In my experience (inside the corporate side of things), the Indian-firm's quality was objectively so much worse than the American-firm's quality that it was laughable to even consider them on important projects.

This mirrors my experience. When I enrolled in a CS masters in 2006, I was repeatedly questioned why I would enter the field when all those jobs would be going to India.

Around that time I was one of 6000 or so laid off and the work sent offshore, at a financial services company. Anyone who'd lived and worked in that period will have seen their or their friends or colleague's jobs at risk for the same.

Exactly for the last 20 years it seems (through ignorance) many non-programmers associate a programmer (even in America) with automatically being Indian.