That's probably not as true within the field as it seems from outside the field.
I have a couple friends that majored in biology in school and then got jobs in biology research labs (NIH, UCSD). Y'know what their day job consists of? Killing rats. And checking rats to see if they're dead yet. And feeding rats so they don't die quite yet. And sitting around playing solitaire while they wait for rats to die.
There are people with Ph.D's seeking these jobs, and every time my friend interviews one, she thinks to herself, "You want this job...why?"
Similarly, my sister is a petroleum geologist, and she needed a master's degree to get her job. You know what she spends most of her time doing? Coloring with crayons. And looking at core sample muck. And poring over maps. And the odd PowerPoint presentation.
I think lots of people think of scientific jobs as being ones where you just sit around and think all day. There is one job with that description: tenured professor. And you need to be tenured, with a bunch of grad students to do the work. Otherwise, your day job consists of dealing with whiny entitled students, lecturing, writing grant proposals, and building experiments in the machine shop.
I know plenty of electrical engineers who are likewise self taught.
Compared to biology, zoology, etc, where the work is by nature far more theoretical, and you're screwed if you never studied it academically.
That being said, there is a side of programming that is theoretical, where having a degree matters - and that's the fundamental, pure math level research that's being done in CS. But let's be honest, the VAST majority of programmers don't have jobs like that.