"Excuse me?" I said to my new boss with barely concealed incredulity.

"You shouldn't thank your people for just doing their jobs. That's what the pay check is for."

When I accepted the assignment, I knew that my boss wasn't known as a "warm and fuzzy" manager. We had very different management styles and philosophies, but this latest directive seemed to cross the line between idiosyncratic and mean.

That was probably the first time I had to think carefully about employee retention. The group that I had just taken over had lost more than 10% of its members in a couple of weeks, and I was afraid that we weren't through yet.

During the dot-com boom, retention was a big topic, but for the past few years, I haven't heard much about it. Now it's back. The job market has improved, and power is shifting away from managers and toward staffs, so retaining employees is back on managers' minds.

Not surprisingly, people seem to be dusting off the same old ideas, many of which I don't find all that instructive. One recent book listed categories of retention strategies that seem utterly typical to me:

  1. Organisational strategies
  2. Orientation and on-boarding
  3. Communication and connection
  4. Career and development
  5. Reward and recognition
  6. Benefits
  7. Capturing employee suggestions

I'm not trying to knock this book, but it's one of many that show how conventional wisdom on this misses the point.

Notice that this is a list of things that a human resources department can do to try to keep employees.

At the root of the problem is that managers and HR departments seem to agree that retention should be handled by HR. HR wants to grapple with this issue and managers don't, so the agreement is convenient for everyone - except the employees and, ultimately, the company.

I don't think that HR is capable of dealing with this - not as a matter of competence, but because of the managerial instruments at its control.

There are two general categories of forces that operate in employee retention: engagement and coercion. Engagement occurs when an employee connects emotionally with his work. Coercion occurs when forces outside the employee encourage either attachment to or disengagement from an employer.

Policy is always about compliance or coercion. HR comes up with policies because some regulatory agency requires it to or because it wants to persuade employees to behave in desired ways. This is not necessarily bad. Incentives can be good, but they are instruments of coercion.

HR creates programs and sets policy. The book's list is a good example of things that HR can do, but I don't usually hear that people leave a company because of bad HR policy: "Gosh, I love my boss, my work and my co-workers, but the incentive plan was so odious that I quit."

Engaged people pay only passing attention to coercion, unless it is truly obnoxious. Disengaged people pay a lot of attention to coercion and usually resent it.

Policy can't enhance engagement, but it can diminish it. External incentives can't make people engage their creativity, but they can tick people off.

Engagement is created on the ground, not at the organisational level. Front-line managers have a lot to do with whether the members of their groups are engaged. Good managers help people stay engaged; bad ones push them toward disengagement.

The goal of policy should be to do no harm and to encourage the development of leadership skills in front-line managers who can make a difference.