I'm going to have to disagree with the article for many reasons: First, I'm a Ph.D student and thus obviously biased. But second, while Ph.Ds can be unwise life decisions for many, it really depends on your field, and the most relevant field to this demographic is computer science. It really doesn't work that way in CS because such a high volume of computer scientists leave academia post-Ph.D into industry. CS industry (and the finance industry) has an insatiable thirst for deeply knowledgeable qualified Ph.Ds. I don't know what the cost benefit is (of spending 4 years in academia vs getting paid high industry salaries) and I'm sure you could make more money going straight into a tech job; but I'm going to assume here that we are maximizing more than just $\sum_{life} income$ here.

PG has a Ph.D; he did fine (yay anecdote). In my various internships around tech companies, there were plenty of senior coders who had Ph.Ds. And if you have a Ph.D in a relevant niche, you're probably going to be headhunted and well sought after. Where else does Wall Street or Google hire top machine learning specialists?

Now a Ph.D in sociology on the other hand... where do you go from there?