With software, I have had this theory for a while that there is somewhat of a law where what was built one year ago or 10 years ago is always cheaper to build today. So this is what helps continue to drive innovation. Have there been any studies on the price of building software always getting cheaper?
I suppose a big exception to this rule would be when there are major players in a particular niche. For example, it would be more expensive to build a competitor to Google now than in 2000. But it still follows that the code it's cheaper to write now because there are so many more tools available to the coder than before.
Writing it now with today's libraries is cheaper than writing the same functionality without the libraries, the way we would have had to do 20 years ago. Libraries keep getting better (for the most part). If they don't, well, you can use the old one or roll your own.
The result of those two paragraphs is that it's always less expensive to write the exact same functionality as it was in the past, because the tools are better. But the problem is that we don't want to write the exact same functionality. We want to add this year's set of must-have features, so the scope is bigger than it was last year...