the failure option

In programming circles, when talk gets around to the fate of large projects, or even small projects, we wonder at the number of projects that fail. Fail as in spectacular, won’t-even-run failure. Fail as in wasn’t what the users wanted, fail as in late, fail as in over budget, fail as in buggy. And so forth. Unlike happy marriages there more ways to fail than all the books in the Library of Congress.
BUT is project failure symptomatic of programming? Or a problem specific to programming projects? I went to the library and pulled down a book on business management. Found the section on managing projects. What did that textbook have to say about project failures. Nothing.
Hm. Now, I have zip business sense, but I had heard, just a rumor mind, that business must innovate or die. Which sounds like projects all over the place to me. And they don’t fail? Not even a concern?
I looked at the another book. I probably write the names down, but I didn’t. And this book had a chart showing failure rates for various kinds of projects. I couldn’t really figure out the chart but it looked like failure occurs quite often.
So are software project failures really different from any other kind of project? Maybe not. Hard to tell. But I suspect not. (It’s convenient for me to think that of course.) It doesn’t make failure a success. But it should give us some perspective. And maybe point us to look at management issues and techniques and not just look inward at our own failures.
Any thoughts?