Less cynically, it also increases the odds of finding good employees. A "bottom 5%" worker is statistically less likely to suddenly rise to the top of the pack than a new hire is to produce at that level initially. And, more cynically, that remains true even if there's significant measurement error in identifying that 5%.

That's true if the performance ratings are cost-free and fairly accurate. A company where firing is based on misconduct and promotion is based on achievement, rather than basing both on quarterly reviews, might actually have a better chance of finding good employees, and perhaps more importantly, retaining them.

Why would companies prefer cheap insurance over productivity? If they want to pay less, they can just pay less. People would rather get paid less than get fired.

Productivity is hard to measure. Insurance cost is easy to measure.

Is it about cost?

It’s often easier to get that pay raise by hopping companies. You’re then replaced with a new hire who is probably getting more than you did. And needs a few months to get up to speed.

Treating your proven employees well to retain them would cost less.

Big companies are driven by the mean and dollars. The short term numbers work out, and long term productivity issues are addressed by acquisition.