Not entirely sure what you're trying to say, but just because someone is unhappy doesn't make them unproductive - certainly and quite likely less productive - but not unproductive (and almost certainly productive enough to be worth more than their cost). While the Economist tries to point some of the blame on measurements on performance, I would make the argument that it's the wrong measures that are the problem. After all, churn/attrition obviously greatly affects productivity/performance as can happiness.

I suspect employers have and do get better at managing people as it's always the best ones who are able to (and often do try to) find work elsewhere in good times and bad. For the employers who don't change, they're left with the results of adverse selection.