Bryan Caplan  

Why They Haven't Been Fired

Henderson and Gochenour on "Pr... Some Observations About Firing...
"Why haven't they been fired?"  Lots of great, non-dogmatic answers in the comments.  Everyone sees a lot of deadwood, though I'm puzzled by the repeated suggestion that "The next person might be worse."  If that's the concern, why not just eliminate the position entirely?  Most respondents think there's rapid employer learning combined with failure to act on that knowledge. 

Why don't employers act on their realization that someone costs more than they worth?  Respondents mention lawsuits, non-profit incentives, and firing aversion.  A few gems:


In my division of a large corporation, I'd say the fraction may be 20-30%.

I think they don't get fired because:
1. Managers don't like firing people all that much - it's a stressful process.
2. If the people have been around awhile, and don't do anything particularly offensive, it's hard to justify firing unless there's a severe budget cut.


5. While it's easier to fire when there's a management turnover, they don't reliably know who the overpaid players really are.
There are numerous people at my job who are a net loss for the organization. I don't mean just overpaid, but rather that any company would be smart, if put in such a situation, to pay them to not to work there. These people generally have both a low IQ on top of an Axis II personality disorder (e.g., borderline personality disorder of antisocial personality traits). But since we're paid by taxpayers (no competition and no bottom line), there's not much of an incentive for their bosses to undergo the long and difficult process (required by union contract) of firing them. So they don't.
1. I work in a very small business (there are only six of us), and I would say that the only person not being paid paid 125% or more of their true marginal product would be our intern that is not being paid at all.
2. They are all personally identifiable, myself included (although not the incompetent part).
3. I think that my boss has failed to fire his overpaid/ incompetent employees because he is in denial. I don't think he wants to admit his current financial predicament. Handing out bi-yearly raises to his employees is not helping his situation either. He feels attached to all of his staff and refers to us a his only family. Frankly, I don't think the problem is having an overpaid staff but having an overpaid staff while business is slow. He is suffering because he is trying to accommodate all of his employees, while the best interest of the business is set on the back burner.
SB's my favorite:

I'd guess 30%. I'd bet I can identify 60% offhand, possibly up to 75% if I really thought about it.

I am one of the bosses in a small business who makes hiring/firing decisions. People seem more likely to cover for weak employees than tattle. For employees I don't observe directly, I usually hear about issues only when their direct supervisor is about to commit hari kari. As to why we don't can most of the ones we know about:

1. Loyalty. The CEO has personal loyalty to some long term stinkers. To some extent, the CEO may acknowledge the lack of value but prize loyalty for its own sake, and I get that (kind of). But people also don't really see the failings they have become used to in long-termers.

2. Customer familiarity...

3. Neoptism. Before I came on board some hiring of the children of good employees happened, with predictable results.

4. Legal headaches. We were sued for religious discrimination by the employee we fired for trying to beat up another employee in the parking lot "to defend Jesus" and for racial discrimination for firing an employee who was stealing cash and inventory by forging customers signatures for purchases and returns (hint: that leaves a paper trail)... Don't get me started on the delicacies of dealing with habitual workers comp filers, or employees with medical problems that make them unreliable.

5. Unemployment insurance. You can't really fire someone for being ZMP and call it "for cause." (Well, you can try, but the argument has to be ... carefully crafted.) So, you have to continue to pay for the privilege of canning them. Since the wage/value ratio is why you are firing them, this can mess it up.


8. No one wants to be a s#*+. More relevant in a small business like this (in a large company I'd imagine managerial ignorance is a bigger factor), but, hey, it sucks to fire people, even people who need firing.

Over all, firing people is a last resort. In all honesty we try to get them to quit, or we just reduce their hours to the point that their impact on the business is minimized.

As a couple of comments note, the financial and psychological costs of firing people if they don't work out makes signaling important even when there's rapid employer learning.  By the time you know hiring a person was a mistake, you already feel sorry for him.

COMMENTS (36 to date)
Peter writes:

'though I'm puzzled by the repeated suggestion that "The next person might be worse." If that's the concern, why not just eliminate the position entirely?'

1: Sometimes 5% productivity is better than 0%.

2: You can't (organized labor in some markets, institutional resistance in others).

You being in academia should easily realize how hard it would be for the Dean to eliminate a tenured position even if that position was unnecessary and empty (let's say the professor just retired). Table of organization changes are bureaucratic nightmare, efficiency doesn’t factor in.

John Jenkins writes:

Two other factors that might be present (1) arrogance; and (2) sunk cost fallacy.

(1) Everyone has some ego and being given a position of power only stokes that. Sort of like the person who dates a series of real losers with the belief that the person can be the one to change each of them, a manager might believe that the manager can really effect a change in the underperforming employee and the only way to show that is to let the person underperform first. The fear might be that the next person is better and won't let the manager "showcase" the manager's skills.

(2) If you have already hired and begun training someone, you might believe that all of that effort will have been wasted if you just fire the person and you don't want to be responsible for that.

blink writes:

Does this imply that a large fraction of employees are underpaid as well? Yes, there are struggling companies like Emily's, but most companies are earning at least modest profits. *Some* employees must be producing above their marginal products. Maybe a tournament model applies -- I bet that most employees being identified as obviously overpaid are long-tenured types who once were highly productive and underpaid themselves.

Joe Cushing writes:

I read one of Jack Welch's books. If memory serves me, GE makes it a policy to fire something like 25% of their staff every year. It sounds like that's the right number, based on these posts.

Joe Cushing writes:

Wikipedia corrects me. It's the bottom 10% and it's only managers.

Scott writes:

A manager confided in me once that he wished he could fire X. He said that if he fired him the big boss would not let him replace him and would cut his budget. Sort of like the logic that "if you can get by without X then you really don't need anyone at all". So having someone do a marginal job, miss work often due to illness, etc. was better than having no one.

jseliger writes:

1. Managers don't like firing people all that much - it's a stressful process.

This is interesting to me because it maps almost perfectly to one of the main reasons professors and teachers don't give honest grades and indulge grade inflation. Ditto for this comment:

. . . since we're paid by taxpayers (no competition and no bottom line), there's not much of an incentive for their bosses to undergo the long and difficult process (required by union contract) of firing them. So they don't.

There's little incentive for honest grading here, either: the professor isn't hurt, or hurt much, by giving a particular individual a higher grade than he deserves, much like a manager in a company of, say, 5,000 people isn't doing much harm to the company by keeping one unworthy individual on.

Cumulative effects, however, are much more problematic, but there's a gap between individual action and the whole.

Elvin writes:

At my old firm, they would order each department to get rid of one or two people every 18 months or so. This happened even in the good times and the cuts were larger in the bad times.

It certainly gave local management the flexibility to can the least productive workers. If you were a weak performer, you could get by for a couple of years, but eventually one of these mini-RIFs would hit, and you'd be gone.

I thought it was a great policy. You could take a chance on someone, knowing that in the worst case, they'd be gone in two years.

Nat writes:

Companies do not want to fire people based on the legal issues. They don’t want to spend the time or the money. However, they don’t realize that they are already spending the time and money based on the non performance of that employee(s). If you look at the non value of that employee(s), perhaps they are sitting around, hardly working, surfing on the net, chatting and/or disrupting other areas or co-workers. Don’t know how to do the work and will not ask so they just do it wrong. Paying this non value employee along with other company benefits (health care, welfare, time off and retirement just to name a few) can be discouraging for valued employees who are actually doing the job. Many companies who do not trim down the non value employee(s) will eventually be seen in the company financials. Otherwise most of the valued employees that do work and produce will soon or later leave that company.

Peter writes:


That is a true statement in two different contexts:

1: Definitely true in non-bargaining unit tenure arrangements though sadly the length of that preliminary non-tenure period is shorter than it should be, i.e. slave away for ten years and then coast for twenty (or in the Federal system, slave away for three years and then sleep at work for the next 27). I’m not faulting the tenure system and can see true value of it in concept, I’m faulting it's extremely poor implementation.

2: HR won’t pay people what they are worth comparative to their peers. If you are hiring five budget analysts you have salary range of X to Y with a deviation within that range allowed for the whims of the hiring agent after which, regardless of performance, you are going to get cost of living raises for the rest of your career and maybe, if you’re lucky, an extra percent or two for being a high performer. Organizations, for a variety of reasons (including you making more than your boss), aren’t going to fire the three non-performing budget analysts, double the highly productive fourths, and keep the marginal fifths the same even though in the end it’s a cost saving for the same work. Sure you see this in hypercompetitive small firms but once you reach a certain size you simply aren’t going to compensate your performers in a fair manner comparative to their non-performing peers. Just imagine for a second if your five budget analysts had to compete for a set combined wage pool every year and your pay was one hundred percent performance based with the salary floor being zero for a year’s worth of work. Even if the ceiling was the entire pot, good luck hiring for that.

Matt writes:

I don't want to be the only idiot who doesn't get it, but what exactly is "true marginal output" in this regard? If a worker is fired and it hurts company moral or affects productivity in some other way, would that be reflected in their "true marginal productivity?"

I'm also confused by all the finger-pointing at managers. If managers are too incompetent to fire employees that are essentially costing their company's money, why aren't they in turn fired?

I'm currently reading, "The Power of Habit," so I want to say that the reason that managers don't fire more is because getting managers into the habit of firing is not productive.

Very interesting question though...

AMW writes:

So the next logical question comes from SB's last sentence:

In all honesty we try to get them to quit, or we just reduce their hours to the point that their impact on the business is minimized.

Why don't more employers just give the lousy employees crappy raises, unpleasant tasks and low status job titles?

Peter writes:


The problem is it doesn't work, plenty of overpaid non-performing employees are perfectly content collecting a pay check complete with crappy raises, unpleasant tasks they don't perform on anyways, and irrelevant job titles.

I work with two guys that are making $83K who literally sleep all day at work, aren't even given menial tasks because they don't perform them, in general are shunned by the entire office, and haven't gotten a raise or bonus other than the organizational mandatory minimum cost of living increase in fifteen years and guess what, they are perfectly content. Both know that they couldn't even get a job making $25K so why would they quit this one ever? This situation is not unique to me or my organization, I have seen in in the private sector also (especially once the organizations get large enough). What people on here seem to be missing is plenty of non-performers are actually fully aware of this and don't care or even actively court said behavior; people seem to be giving them the benefit of the doubt that they are just unmotivated, poorly managed, unskilled, etc etc when that's not usually true. They are effectively welfare moms of the corporate world and usually the best way to get rid of them is to literally promote them elsewhere after giving them shining reviews; beware of gift horses from other organization units :)

Andy writes:

Some other reasons:

1. Firing people hurts morale. Very productive employees are often irrationally worried about being fired (probably because their look at their work critically, which is what makes them effective). Firing people, even those perceived as not as competent, may make them unhappy.

2. Firing people encourages people to "play games". Most employees don't really want to spend all their time being "political". If the potential payoff is just a slightly higher raise then many employees will not bother. But if they perceive their job as being threatened, they might spend more time on this. Thus you may hurt overall productivity.

3. Some employees may improve. Especially in fields where experience counts for a lot, they could move from being below marginal cost to above marginal cost. If they aren't *too* much below, then it may be worth it to take a chance, especially in fields where it's difficult to find good employees (i.e., most jobs).

4. It's hard to measure "marginal value". Sure you might know that an employee is in the bottom 10% of performers, but it's hard to translate that into them actually being worth less than their marginal cost. It's certainly possible that every employee produces some marginal benefit. Perhaps managers are often reluctant to admit that they made a mistake, and thus justify keeping the least good employees by the argument that they still provide some value.

David writes:

somebody hired the dead weight; why not fire that person too...or why don't we train both morons...obviously someone saw something positive when the candidate was hired. either the hiring manager has no business doing the hiring or the candidate has not been adequately trained to do the job well.

Oliver writes:

I get the feeling there must be stable equilibria in between policies "fire as soon as a worker demonstrates non-temporary negative net product" and "fire only when significant socially acceptable reasons to fire". I.e., there must be optimisation still to be done that preserves signalling.

Writing this comment has made me realise how poorly prepared I am to answer it.

unproductive writes:

many commenters here seem to *care* about unproductive workers in their organisation.
I think the truth is that in a large organisation most people don't care.

Peter's $83k dead weight colleagues are good example. they are unproductive, they don't care much, their bosses don't care much, life goes on.

If they were actively disruptive and unplaeasnt to have around their boss would probably do something otherwise, why bother? what difference does it make?

John Fembup writes:

I may have missed your formula that determines objectively whether a given employee's employment cost is at least 25% greater than their true marginal product. That might be the reason why my comments seem to be a few percentage points over.

Meanwhile I'm stumped about how the true marginal product might even be measured for a physician, for instance, or a cellist in the symphony, or an economist.

However, as an alternative to measuring marginal product, the GE approach seems to me to have the advantage of transparency. It is arguably transparent to identify the bottom 10% and arguably transparent to have a policy of removing the bottom 10% every year. I say "arguably" because incompetent managers make poor, subjective employment judgements. These same managers have the incentive to hide their shortcomings and have the means and company standing to blame the results of their failures on others. In fact, this latter observation even applies in politics.

Dan Carroll writes:

There are several issues with identifying non-productive employees:

1. Sometimes employees are non-productive because being productive does nothing for them. Raises are automatic or their bosses are clueless. Or their extra work is discarded or diluted by others.

2. While some ZMP employees may be obvious, most are not obvious. And it is not obvious whether the problem is the employee, the job, or the oversight.

In my industry (finance), it is not unusual to get rid of employees deemed to be non-productive. Unfortunately, there appears to be little discrimination between those that should go and those that should be given raises. Instead, it is based more on personality, clashes of ego, and other irrelevant factors (like who is boss is or what school he went to).

It is easy to forget that organizations are made up of people, and people have agendas that are not related to the profitability of the organization.

unproductive writes:

in a large company almost noone is bound up with, or often even particularly interested in the overall profitability of the organisation.

overall profitability is bit like the weather, or the climate. It changes, it's unpredictable, sometimes there is a storm and bad things happen, sometimes its sunny and all is well.

you can often somewhat predict it -- but not always
you can, often, usefully prepare for it

but you can't affect it, and most days you just ignore the weather and get on with your life.

Eva writes:

What about the effect on the morale of other workers?

There are situations where getting rid of someone could improve the atmosphere, but I think the reverse would be more the norm....

JCE writes:

and how exactly are we to determine their (o anybody else's) marginal product. there is no way.

even more so in the context of a company or firm. wasn't that the point of alchian and demsetz's paper "Production, Information Costs and Economic Organizations," ???

Glen Smith writes:


The unproductive worker may not have been an unproductive worker when he/she was hired. This worker may not have been able to adapt to business changes or the company may not have adjusted to new realities in their industry. In some fields (such as operations research), almost all highly skilled employees will go through periods of negative marginal product and since the hiring costs could be prohibitive, it is often better to keep them on staff (especially if they have a rare skill set). Finally, not being the "bad guy" may have high value to me so often the employee who has seems to have a true marginal product of less than 0, may in fact have a true marginal product greater than 0 (and conversely, an employee who appears to have a positive marginal product may find his/her true marginal product is less than he/she thinks).

Jason White writes:

Several people commented above about companies that fire x% per year. I have several friends that work as management in companies that have that sort of policy. It has the perverse effect of increasing demand for poor workers, so that you have someone you can sacrifice. They tell me managers actively fight over poor performers as the deadline for firing looms. Also, they tell me that if they fire someone, they are just going to get them back in 2-3 years, as that employee goes from company A->B->C->A (all of which follow similar firing policies.)

Jeremy Skog writes:

There are many externalities that need to be considered that can make it sensible to keep someone on even if they're below their MPL. Just a few:

Firings can lower/morale and increase stress. The net effect summed across all employees can be greater than the initial gain from dropping one employee.

Recruitment costs can be high, especially in skilled positions. Getting 80% of someone's MPL can be worth more than having the position vacant for 6 months or paying a year's salary in time/effort/money to recruit a replacement.

Depending on the degree of specialization there can be a lot of training involved so that a new employee might take some time even getting up to the productivity of the "low MPL" worker, let alone surpassing it.

SB writes:

Re: firings affecting morale - I've seen numerous cases (as a boss and as an employee) where the other employees are encouraged by a waste of space getting fired. It may reassure them that someone is paying attention, hand on the tiller and all that.

Of course, the fired person has to be generally known as dead weight for that to be the case.

Re: alternatives could be worse - I'm well aware that, when you know someone's not working out, you are better of trying anyone else, because at least a random guy off the street has some chance of working out, which is a better chance than what you have. But that kind of (blindingly simple) logic requires a certain amount of effort to overcome typical human status quo bias. Inertia is easier, and risk aversion is a very easy excuse.

bxg writes:

And sometimes this: the manager actually assigns some utility to the well-being of this marginal worker. The manager has some other objective function - maybe involving compassion - that says shareholders' interests ar not the entirety of her objective function. And so a worker who is doing good work, just not enough good work to cover his salary, may still be a keeper under a utility function where his well-being actually enters the picture somewhat.

After all, why does the CEO get paid so many tens of millions? He seems to be a moron whose best attribute is belonging to the right golf club. (You don't have to agree that CEO salaries are ever in the tiniest respect economically inefficient; the point is that some managers might perceive this is so). And maybe myself (the manager) think I am overpaid - I don't think I could justify my salary relative to my marginal contribution (maybe because it's really unjustifiable; maybe because I'm not good at the economics of marginality; maybe I'm just insecure). I the manager look around, and see people acting in the interest of themselves only and/or their peers? Yet I should be the lone wolf with a laser-like focus on the shareholders? Why? Let the CEO lead us there first and I might listen.

Skeptical writes:

Wow. This should be damn near self evident to anyone who's worked in corporate America for any amount of time. But Mr. Kaplan lives in his bubble, so I'll spell it out for him.

In any company large enough to have an HR and legal department, managers don't fire people, they document. They have to document every step in the process: what the standard is, how the employee didn't meet it, when and how they notified the employee about not meeting the standard, what corrective process they worked out with the employee, how the employee didn't follow the corrective process, how the substandard performance kept occurring.

This isn't a weeks long process, it's usually three to six months. If the deadwood is in a protected class it's even worse.

Heard the term 'hiring freeze'? That means the manager can't replace the trimmed deadwood, or they have to go through the justification process much higher up the corporate ladder. And as others have stated, there's no guarantee the replacement will be any better.

The deadwood are quite aware of this process, and they'll game it. Getting put on corrective action is no big deal. Slack for a couple of months, then start working again and go off of it.

Exceptions: Racial or sexual discrimination. Sexual harassment. Workplace violence or certain things that maybe perceived as violent. Not meeting targets for commission based jobs.

This isn't meant as an ad-hominen attack, but this post shows you're the very caricature of the ivory tower academic. You need to get out of your bubble and see that marginal productivity isn't the end all, be all in the real world.

Eric Rasmusen writes:


you haven't really answered the question, you've just rephrased it to, "Why do private companies have HR departments that impose stupid rule that make it hard to fire people?"

You might answer that with, "Because HR people don't care about profit," but then we must ask why the CEO gives them the power to hurt the company.

That's why economists are puzzled.

Right Wing-nut writes:


If you sufficiently narrow your field of inquiry, anything is a mystery.

Companies have HR & legal departments to reduce lawsuits. I'll leave it as an exercise to re-read the comments here as to what that means in practice.

Peter writes:

Yep at what Right-wing nut said. The so called "employment at will" ideal that is so often touted and cherished is really a one-sided arrangement where (outside very limited cases) the employee can quit without regard to the company but the company is hamstrung on the flipside. The only places I have ever seen true employment at will (from a company perspective) is in small businesses beneath the EEO thresholds. This not to say people are being fired for discriminatory reasons but that the employers can simply fire people without fear of superficial lawsuits claiming so.

Skeptical writes:

[Comment removed pending confirmation of email address and for rudeness. Email the to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Anthony writes:

There are some natural experiments in this question just waiting to be researched:

1) Difficulty of accurately measuring productivity: Some consulting firms bill their employees hourly - one can measure an employee's productivity by how many hours they actually do bill, after adjustments for work that can't be billed, etc. Are those firms more likely to lay off poor performers than firms where it's much harder to determine any individual employee's individual production?

2) Difficulty/cost in replacing/training skilled workers - look at firms with highly-skilled or specialized workers versus firms with relatively low-skilled or easily trainable workers.

3) Documentation requirements: look at union versus non-union firms in the same field with the same sorts of workers. Look at larger firms which have documentation policies versus smaller firms.

4) EEOC issues: look at firms whose workforce is almost entirely white versus those (of similar size) whose workforce is racially mixed. Bonus points for finding firms whose front-lien workforce is almost entirely Hispanic or black.

5) Agent/principal problem (managers don't care enough about company's bottom line, or are empire-building, etc.) Look at firms where the managers with firing authority also have significant ownership stakes versus similarly-sized ones where they don't.

mark writes:

Firing people is disruptive. In many contexts, its short-term marginal cost likely exceeds its short-term marginal benefit. I don't mean severance, I mean the selection process, the legal b.s., the stress of informing people (both the fired and the ones who stay behind), the work needed to restructure assignments and functions, etc. It's only when you get into real cash crunches that the MB > MC. That is also when you can hand out a story that minimizes the psychic turmoil, i.e., you can say, it's not that I have a low opinion of you, I have no choice.

Like government, another organization, absent a crisis, people prefer not to engage in unpleasant tasks to fix things.

Bryan Willman writes:

The documentation/legal hassles and even "fire x%" comments hint at something else.

Firing someone is by definition a management activity.

Management resource (time, attention, energy, etc) is finite.

So for some part of the time, it will be a choice between "do I spend 100 hours over the next 6 months firing Pat loser, OR so I spend the same 100 hours developing an important new account?"

Unless Pat loser is truly awful, firing them will be a lower priority than most anything else.

If the org has a lot of losers, there may be a "fire losers budget" in terms of management time, litigation risk, etc., so they fire the worst N losers every year. And no more, because they cannot *afford* to.

Kevin writes:

What about the classic screening model? Employers commit to pay employees less than their marginal product in the early years, but then pay them more than their marginal product in later years.

The only employees who will accept this deal are those who are patient and confident in their ability to prove themselves as the productive type. As long as management can commit to paying more than a worker's marginal product later on (call it 'loyalty' or 'entrenchment') then we can hit a separating equilibrium.

Comments for this entry have been closed
Return to top