Thursday, June 21, 2012

Who would you let go?

More notes on an occasional topic on this blog: firing. First, the Navy, as is traditional, let another ship captain go and told everyone about it (via Tom Ricks). Not for zipper problems, but for not handling his ship correctly:
The commander of the big-deck amphibious assault ship Essex was fired Monday due to a “loss of confidence in his ability to command,” a Navy spokeswoman said Tuesday. 
Capt. Chuck Litchfield was commanding the Essex on May 16 when it collided with the replenishment oiler Yukon a day before both ships returned to San Diego. “There were a number of factors that contributed to the collision with Yukon,” Reese said. “Part of it began with the loss of rudder control. There was a breakdown in command and control, in bridge resource management and in communication between the two ships. … All those factors contributed to the collision.” 
The problems “essentially began with the partial loss of rudder control,” she said.
When did you last hear of a pharma or chemical industry manager being fired for job-related performance issues?  While the life-and-death stakes are not as large, the financial costs to mismanaging a large project in industry is probably similar (or within an order of magnitude) to fixing "parts of the starboard elevator, lifeboats and catwalks... the flight deck and davits."

On a similar note, you are presented with 4 hypothetical people, one of whom you need to fire. Here are their profiles*:

Who would you dismiss? Via Bryan Caplan, here's how a study with a similar example worked out (I moved the years ahead to 2012, kept the ages and performance evals the same, and converted from pounds to dollars):
Who gets fired?  Almost half the respondents make what sounds like the profit-maximizing decision - firing #3.  But #2 is almost as popular, and almost 15% fire the older but excellent performer.  (See the bottom row of Table 2 for details).  More strikingly, though, answers vary radically by country... England fits the standard caricature of the "greedy" Anglo-American business model: over three-quarters fire #3.  Germany is at the other extreme.  Almost three-quarters fire worker #2, with worker #4 a distant runner-up.  Spain is closest to England, with France and Italy about midway between England and Germany.  In Italy, over 20% fire the excellent #4.
Click on the link to find out why. (Culture and economics play a role in how these decisions are made, unsurprisingly.) Personally, I would have gone with ol' #3 myself, assuming that the performance evaluations are fair and believable. Hard to say, though.

Readers, what would you have done?


  1. If I was #1 I'd quit! But I went for #3 too - too expensive to justify middle of the road performance.

  2. Corporate HR DroneJune 21, 2012 at 11:58 AM

    Fire them all and outsource their jobs.

  3. The Navy has fired some crazy number of commanders this year, approaching 20 I believe, for a variety of reasons.

    Let's face it, one of the main things as the commander of a vessel is to keep it from striking other vessels. If your people can't handle that task, then you haven't done your job as their CO.

    Anyone considering firing #5 should know that unless you fire #1 or 2 as well, chances are you'll get sued for age discrimination. When our layoffs happened a few years ago there was a really nice spread of age/ethnicity/sex so that no-one could come back and claim discrimination. Unfortunately that also means that good people lost their jobs simply because they needed a certain demographic covered.

    1. There are some things I might consider changing about the Navy (length of stay for commanding officers, etc., joint-ness requirements, etc.), but their tradition of holding COs accountable isn't one of them.

      I think the weird "HR wants a overly smooth bell curve" of layoffs is SO strange and unfair. Thanks, lawyers!

  4. I also went for #3. I really can't understand why you'd fire #2...aim for retraining, or job shift.

  5. CJ:

    Speaking from experience here: US Naval officers have it drilled into their skulls *from day #1* that any collision (ship, sand bar, etc.) at sea is career ending. And the end is swift and sure.


    1. You may not even be the de facto responsible individual (e.g. the watch officer) whose action(s) or lack thereof led to the collision. An event like this, the sh*t flows uphill to the ship's ranking officer.

  6. Unstable IsotopeJune 21, 2012 at 3:51 PM

    I'd fire #3. I'd probably fire #4 over #2 for second choice.

  7. Depends on how much you can trust those evaluations - mine varied quite a bit depending on which boss I had at the time!

  8. @907am
    Bell curves are real. Wish companies would base decisions on performance.

    From what I've observed:
    Protected classes emerge from lawsuits, which cause some discriminatory practices so companies can statistically prove non-discrimination while protecting against lawsuits from high risk demographics.

    Age must be perfect bell curves.
    White male over 40 are protected.
    Minority female under 40 protected.
    Minority male over 40 gone.
    White female under 40 gone.
    White male 35-40 automatically gone.
    Young white male are never or very rarely hired.
    Minority women preferred in hiring.

  9. Fire 2, 3 and 4, hire a nephew, give myself a raise.

  10. fire everyone and hire indian/chinese workers under skill-shortage pretense. pay foreign workers peanuts, claim to have saved companies millions of dollars and help myself to a big end-of-year bonus. Winning!

  11. I agree with anonymous that the ratings are just about useless. My experience has been that the rating depends most on the manager, and not your performance.
    You need to know why you need to fire someone. If it is to save money, then number 4 goes, and you get a nice bump to your bonus for coming in under budget. If they can’t tell you why you need to fire someone, probably fire number 4 hoping it is budget related.

  12. tell #1 how much #2 & #3 make. This will encourage #1 to quit on their own so you don't have to pay severance or unemployment. If luck is with you they will be wildly successful at their next job, which will encourage other strong performers to quit without you ever having to make an unpleasant decision!

    Sure your company will eventually be full of incompetent slackers, but noone gets fired.

  13. In Germany company loyalty still means a lot. There are many family owned engineering firms (~100 workers) where usually the firings are pretty rare. With the new government rules in place that make the worker work part-time with government picking up half the salary, it encourages companies to keep workers during down-times so that they emerge in a good position when the market picks up. Firing #4 is very bad for morale in Germany and you don't want to do it. Plus, with the new government rules, #4 is the most likely to take a time/pay cut for a few years as they are also the most loyal to the company.

    My research on this is from reading German business news.

  14. I guess I find it surprising that #4 would be on anyone's list. If you assume that the performance rating means anything, which I agree can be variable, then you're firing a top worker who doesn't make that much more than an average worker in #3. The cost savings between the two is, for the most part, negligible.

    If the ratings do have meaning, you've just killed morale because they see you firing a top performer for nothing more than the fact that he makes more money, probably due to higher raises due to his performance, not to mention the fact that you've opened yourself up to an age discrimination lawsuit if you work in the US. You could offer him a juiced up early retirement package, but if you fire him he'd have a pretty good shot at coming back with a lawyer, especially after you kept three people who were younger and not as good of a performer.

    1. If I'm a middle manager and my quarterly evaluation depends solely on whether I can reach set goal of 120k in savings it really doesn't matter who do I let go - 1 and 4, or 2 and 3. Both work out as far as I am concerned.