How to calculate the cost of turnover–carefully

By: | August 29, 2019 • 3 min read
Peter Cappelli is HRE’s Talent Management columnist and a fellow of the National Academy of Human Resources. He is the George W. Taylor Professor of Management and director of the Center for Human Resources at The Wharton School of the University of Pennsylvania in Philadelphia. He can be emailed at

Is there a more fundamental metric in human resources than turnover costs? It’s the one measure that (almost) everyone understands and thinks is important. The cost of losing an employee is one of the most basic factors driving investments of all kinds in HR, from more careful hiring to career paths to explicit retention exercises.

People with some history in this industry will remember the field of HR accounting, which basically generated internal accounting-like answers to the costs, and sometimes the benefits, of spending on human resources. There are still plenty of calculators that will generate estimates of the costs of turnover; however, those estimates always seem low to me. One reason is because they often just count the cost of replacing an employee. In other words, they count what can easily be measured.


What we would like to do is know—beyond out-of-pocket hiring and onboarding costs—the effects of someone leaving on the overall operation of the business. That might include the impact of leaving positions unfilled, effects on morale—everything.

Into this question comes a new wave of economists who come at it not by trying to total up separate cost items but by looking to see what happens to overall organization outcomes before and after people leave. One of the first problems they recognize is that it depends a lot on who leaves. When a bad performer is fired, or even when a poor performer quits, we might expect the effects not to be so negative, maybe even to be good. So, how do we sort that out?

One way has been to see what happens when employees die unexpectedly. (And who says economists aren’t fun?) Death seems to hit people randomly, with respect to their job performance, or at least that’s what we think. But those studies are pretty much focused on leaders.

Here’s a new one and a good one, by Peter Kuhn and Lizi Yu at the University of California Santa Barbara. They looked to see what happened to store performance in retail operations when employees left front-line sales jobs. The particular operation they studied had a policy where employees had to give two weeks’ notice before quitting and, for practical purposes, they did not fire or lay off anyone in the period in question.

The punchline? They saw four rounds of negative effects on performance: when an employee first gives notice, just before they depart, right after the employee leaves, and finally, when the new employee is brought on board. Surprisingly, two-thirds of the damage comes before the employee actually leaves: It is not the case, at least in this study, that most of the costs come from being short-staffed after the employee leaves, but no doubt that is at least partly because of the fact that these lower-level jobs get filled reasonably quickly.