Bryan Caplan  

Pick Your Poison

Germany: trading political sta... The Big Bag of Wealth...
Suppose you need a helper, but face two imperfect candidates:

1. A smart person who only cares about himself.
2. A stupid person who only cares about you.

Under what conditions would you prefer #1?  Under what conditions would you prefer #2?  Why?  Please be specific.

Here's my default answer, but I can picture some good exceptions.

HT: Jason Brennan

COMMENTS (34 to date)
Bob Murphy writes:

At first I thought this was a post about the marriage market.

Bryan, In my country, every organization has something to hide. If someone hires a smart person who only cares about himself, I think there are many ways in which this smart person can make the employer say "Sorry".

George writes:

Bryan, I have run a business for many years. I have hired both smart and stupid people, selfish and selfless.

I have to say that you can counteract the selfish employee with rules about what you expect out of them. Their raises and continued employment depend upon their performing their duties and adhering to your rules.

A stupid employee will constantly make extra work for you and, occasionally to frequently, will cost you money with their mistakes.

You can always work with smart. Stupid may include well-intentioned, but you will regret having them.

Mike Gibson writes:

I don't know about conditions, but this questions remind me of Kurt von Hammerstein-Equord's taxonomy of talent. It's a great quote:

"I divide my officers into four groups. There are clever, diligent, stupid, and lazy officers. Usually two characteristics are combined. Some are clever and diligent -- their place is the General Staff. The next lot are stupid and lazy -- they make up 90 percent of every army and are suited to routine duties. Anyone who is both clever and lazy is qualified for the highest leadership duties, because he possesses the intellectual clarity and the composure necessary for difficult decisions. One must beware of anyone who is stupid and diligent -- he must not be entrusted with any responsibility because he will always cause only mischief."

Michael D writes:

In order to clarify my preferences, I needed to add a third person to the mix.

1. A smart person who cares only about himself.
2. A stupid person who cares only about me.
3. A stupid person who cares only about himself.

I suspect that person #3 is less likely to accidentally harm me than person #2.

Therefore, I prefer 3 to 2. And I prefer 1 to 3. Assuming my preferences are transitive, I must prefer 1 over 2.

Joe Teicher writes:

If what I need help with doesn't require intelligence then I'll take the selfless person, because presumably I don't have to pay them. If the task requires intelligence then I guess I'm forced to take and pay for the smart person. This seems very obvious to me. I can't understand any answer that doesn't account for the differential need to incentivize the selfish and selfless person.

Joe Cushing writes:

I'm thinking that I'd prefer the stupid person however there may be times when the stupid person is incapable of what needs to be done. Only then would I consider the smart person. Even if the smart person only cares about him/herself that could be used to get quality work done. Most of us care primarily about ourselves. Compensation plans are often designed to take advantage of this. They are often gamed though, which a smart person would be able to do.

Tom West writes:

> Most of us care primarily about ourselves.

Perhaps, but out self-perception is critical to our happiness, and that self-perception is that we are loyal, and thus do *not* act in ways that directly harm others, even to our self-interest.

At least in my limited experience, the few organizations I've worked for where it was *acknowledged* that everyone was in it for themselves were horrible places to work for everyone. (I was too young and inexperienced to know how productive it was, although I suspect not very, despite relatively decent talent.)

John Fembup writes:

Isn't it a fundamental economic premise that people acting in their own interests will nevertheless produce aggregate results that benefit others around them? So hiring a self-interested person is not necessarily a problem.

So in most instances I'd favor the smart person. Bringing intelligence into an organization is usually a good strategy.

Like you, I can imagine exceptions.

vidyohs Blanco writes:

I would hire the smart person who only cares about himself, with the understanding that he might not be with me a real long time.

However, a smart person can learn that to help himself, by working for me, is only possible by helping me.

Stupid, unlike ignorant, can not learn, so I would be wasting not only my time with hiring a stupid person, but I'd likely put my business in jeopardy as well.

Steve S writes:

I think the answer depends on how you frame the characteristic of "only caring about yourself".

The knee-jerk reaction is to say this person will be selfish and act in Gordon Gekko-like ways to do the least amount of work while extracting the most amount of salary/perks as a part of being your helper. My thought would be that someone interested in himself realizes that his employer has the power to fire him and will work to maximize his value so that he retains his position of helper.

So to answer your questions -

Under what conditions would you prefer 1? When my helper is needed to perform useful/technical tasks.

Under what conditions would you prefer 2? When my "helper" is someone there to perform menial tasks or raise my self-esteem.

I see #2 being the preference less than 1% of the time.

Blade doc writes:

This is the old 4X4 active/lazy, smart/stupid matrix. Active smart can be most helpful (if your motivations align), but active/stupid is the most dangerous whether or not your motivations align. Lazy/smart is probably better than lazy/stupid unless your goals are not aligned at all.

RPLong writes:

Both the smart person and the stupid person will be easy to motivate, but only the smart person will consistently perform well if properly motivated.

Richard Monihan writes:

I'd take the smart person who only cares about himself. Why? Simple. He's more intelligent and knows that, sometimes, making sure I get what I want and need will benefit him tremendously. It's been my habit, in hiring, to always hire someone I perceive to be smarter than me, regardless of his personal view on who he 'cares about'. He could be the most charitable boob in the world, but if he thinks picking up my laundry and getting my car washed is productive, when I need him to run actuarial analyses, well...who's coming out ahead, me (who has to do the work, but has his laundry and clean car) or him (who avoided the real job, but is looking out for me)?

What if the less intelligent person falls on the sword in a meeting where I made an error? Does that benefit me? Sure, he takes the hit, but as his boss, my superiors look at me sideways. Who lets a person make such a big mistake? In that instance, we both lose. Better than the intelligent guy stay quiet and I take the hit. My superiors would be upset about the error, but would respect the willingness to take responsibility for something that I could easily have put off on my co-worker.

The smarter person will realize that his/her behaviors may tend to overwhelmingly benefit themselves, but in the long run their behaviors will ultimately benefit me (unless they use their intelligence to constantly undermine me or throw me under the bus - but that becomes very transparent over time and a truly smart person doesn't do that).

Brian writes:

As others have noted, it depends on what you mean by "only cares about himself." If by that term you mean "stupidly selfish," as in being late for work every day because they want to sleep in, then I doubt anyone would want them. But since you specify that they're smart, I will assume you mean "smartly self-interested." In that case I would prefer #1 in almost all circumstances.

But let me go further. I would prefer "smart and cares only about himself" to "smart and cares only about me." Why? Because anyone who acts on my behalf is likely to be making presumptions about my preferences that simply aren't accurate, and that's of no use to me. Moreover, it likely means they are frequently intruding on my daily activities when I mostly just want to be left alone to act on my own behalf. Frankyl, I don;t need the aggravation. If I want something done that I can't do myself, I'll ask you and pay for it. The self-interested smartie is most likely to respond to my preferences under those circumstances.

Aaron Zierman writes:

Take the smart person in all cases as long as "help" means "production".

Assuming that the smart person is able to produce more than the stupid, this will lead to an overall increase in production (yourself + smart person). [Apply comparative advantage here] Since trading (appropriately) results in mutual benefit, both yourself and the smart selfish person will be better off, thus fulfilling both self-interested desires.

Daublin writes:

One variable to consider is the size of the firm.

In a large firm, a selfish person can engage in turf wars and undermine other smart employees. This lowers productivity in the short term, easily making their net effect negative. This also harms morale for anyone who figures out what is going on, thus sucking more people into the selfish camp.

In a 2-person operation, at the other extreme, there's not so much harm you have to worry about. The selfie can just shift a greater percentage of the proceeds from you to them, but if they are smart enough to increase the size of the pie, they're still worth having.

Eric writes:

All right, I'll bite. Everyone is approaching this problem as though you are hiring an employee and paying the market rate, as well as creating an incentive structure. I.e. do you want to "pay" for smart. This is one possible story Caplan is after.

Another story is: I know perfectly well a stupid person who cares only about me. Me, of course. And smart people who care about themselves. Financial advisors. (Actually, I think I'm smarter and better at finance than most financial advisors but stay with me). Would I rather manage my own finances, or outsource them at an appropriate incentive-compatible fee structure. I (and most people) would choose self-management. That is, stupid but self-interested. Govt is an extreme example of the best and the brightest, but not interested in me as much as I am.

MikeDC writes:

Intelligence and intent dominated to the point of irrelevance by the importance of political power and control.

I can fire the helper and hire another at any time. I'm stuck with the ruler, and the best solution to either the problem of "too smart" or "too stupid" is to give the ruler less control and give me more.

I can't think of any systematic reason, in a political context, why I'd chose the well-intentioned but stupid over the smart and selfish. The link between self-interest and mutual benefit is so much weaker in politics than it is in market economics that I'd rather not rely on it at all.

A smart, self-interested politician need only build a winning coalition to be successful. Worse, they can build a coalition by creating supplicants beholden to policy whether the policy is all that good or not.

Brian writes:

Hopefully I can design a contract that gives the smart person incentive to do what's in my best interest.

I can't design a contract to help a stupid person make smart decisions.

Yancey Ward writes:

What kind are implementing Obamacare?

Chris H writes:

@Joe Teicher

I think that's an interesting answer, and it's certainly shaping how I approach this.

A stupid, selfless person should be cheap to use due both to his disposition, and the presumed lack of skills. The smart selfish person will be relatively expensive due to both skills and outlook. This makes me think that in equilibrium I would trend towards indifference between the two.

This is especially true given that the low-skill person is specified selfless, and thus unlikely to have the morale issues that normally come up with pay differentials.

Glen Smith writes:

It really depends on what I need the person to do. In most cases, the stupid person who cares only about me while the intelligent one who cares only about himself is usually the most dangerous. If I know the person is stupid, I can watch him and be careful of what I ask him to do. The intelligent one (especially if he is not lazy)is likely to be more dangerous. You could implement rules but intelligent people are very good at "following the letter of the law but not its spirit" (along with either the ability to justify their actions to themselves or not need any personal justification).

MingoV writes:

I'm assuming that I can provide some form of compensation for the helper. I also assume that a position of 'helper' is not permanent. If those assumptions are true, then I would choose the smart egoist. The smart person has the potential to be the best helper. All I have to do is create work conditions and provide compensation that meet the helper's egoistic needs. If that isn't possible, I'll search for a different smart person.

Daniel Fountain writes:

#1 By paying him what he wants his interests hopefully then shift entirely to completing what I want so long as his expectations remain static for the life of the project. Of course if I can't meet or exceed such expectations he will most likely have a detrimental attitude toward his work, or focus more on his opportunity costs (i.e. other jobs). Lets give P(D1) to meeting his demands and assume P(T1) = P(D1) since as long as I meet his demands he will complete the task.

#2 This individual doesn't care about payment so long as I tell him he's doing something positive for me. So as long as he is able to perform the task I set out for him within the time I need he will be my default choice. Lets give probability P(T2) to him completing the job since T2 is independent of D2 (which is always 1).

At this point there are two systems with P(T1) = P(D1) as the probability that #1 will complete the job and P(T2) for #2. Now with P(T1) = P(T2) it would be foolish to pay #1 since #2 costs nothing. Thus there is a system of equations where the cost of P(T1) is (assuming linearity) k*P(T1) and the value of the task is V. So when P(T1) satisfies P(T1)*V - k*P(T1) = P(T2)*V is the point past which paying #1 will reap better benefits than paying #2. Note that k will be inversely related to #1's non-monetary interest in the task.

This could be further analyzed by giving P(T1) or P(T2) some time-dependent function and ceiling, some minimum necessary probability of finishing the task, etc. etc. Of course in the real world people would have to make guesses at the probabilities but over time I'm betting their guesses would become more and more accurate.

LD Bottorff writes:

I understand people who are smart and only care about themselves. I'm inclined to hire someone I understand.

Stupid people can be manipulated, but they are still unpredictable. If I could predict when and how they are going to screw up, I could work with that. Unfortunately, I'm not that smart.

hana writes:

I would never pick based on only these two characteristics.

However, if you changed them to speaking Japanese and dumb, vs. not speaking Japanese and bright, depending on the circumstances I might have some trouble. (sic)

The point is that all skills exist along a scale. All jobs exist along a scale of skill requirements. Employers, unshockingly, pick employees that fit within the parameters of the scale. Expectations by employers of employee growth enable them to expand the requirement windows.

Jerry Ward writes:

Hire the smart person under an arrangement such that his self-interest coincides with mine.

Ed writes:

It's interesting that you found no significant role for income in determining voters' behaviour. I wonder if this result would have been the same had you analysed voting behaviour in other countries. I think that income is much more correlated with party affiliation in the UK (and most of Europe?) than in the USA. This might be because British parties' differences are almost all economic: we don't have any equivalent to your culture wars.

English Professor writes:

1. A smart person who only cares about himself.
I prefer this person if I have control over his compensation. I reward him well if he takes care of me as I want.

2. A stupid person who only cares about you.
I prefer this person if I have control over his actions.

Lars P writes:

I think it depends on how observable the quality of the employee's work is.

If I can easily see the quality of the work, perhaps in the number items made, I can reward the smart person for the value produced, make our interests align, and thus remove the problem.

If the produced quality is hard to observe, things are much dicier. I'd either have to hire the dumb but trustworthy person or no one. The smart cynic would easily figure out how to get paid for doing nothing.

Job types where this might apply are teacher, agent, therapist, consultant. Jobs where you're selling an expertise that the customer can't understand, and the customer therefore have to take the results largely "on faith".

Debashish Ghosh writes:

Prof Caplan, it would really nice if you could please also post your default answer in a more succinct form..

Greg Jaxon writes:

The conditions under which I'd prefer #1 are any where the nature of the help involves taking risks for significant reward and forming a partnership with #1 will increase both of our expected outcomes.
E.g., I am an annuitant and she is an inventor.

The conditions under which #2 are preferable is where the nature of the help is that it unlocks value mostly to me and involves few risks.
E.g., I am becoming demented and a caring relative is willing to move in to help.

Martin writes:

Isn't this just the bias/variance trade-off? Smart is biased but low variance. Dumb is no bias but high variance.

Comments for this entry have been closed
Return to top