When Incentives Backfire

In business, certain ideas have come to be treated as received wisdom.

One of them is “align goals and incentives.”

It sounds dirt-simple. If you want to encourage or incent people to do certain things rather than other things, then align their incentives with those things. Reward them for doing the desirable, punish (or do not reward) for doing the undesirable.

Praise the child for helping, discipline the child for misbehaving. Say “bad dog” for jumping on the sofa, say “good dog” and give a treat for heeling.

Increase the commission on the profitable product, decrease it on the lower-margin ones. Give the CEO stock options to incent him or her to increase the stock price. And so forth.

But this idea has an exploding-cigar component to it. In fact, it can be downright destructive.

In the recent Republican Presidential debates, one candidate suggested, with an “ain’t it obvious” kind of tone, that part of the answer to the US health care problem is to incent providers based on outcomes. We should pay doctors and insurance companies for improving people’s health, then they’ll work to improve patients’ health, thereby cutting health costs.

I mean, ain’t it obvious?

Look a little closer. It suggests that, as a doctor, the most attractive patients will be overweight smokers—because I can quickly improve their health. And I will work hard to get them to diet and quit smoking quickly, because I get paid more for showing fast results.

On one level, this is very good. It’s a form of social triage—focus on the highest improvement rates possible. Obesity and smoking are major health problems. What’s the problem?

The trade-off is e subtle shift in motivation for doctors. If everything goes through the money filter, where’s the motivation that keeps interns doing absurd things to their circadian rhythms for so long? What happens to bedside manner?

What happens to any sense of the purpose of medicine—which, since roughly Hippocrates, has been largely based on healing people? It gets replaced by the same motivation that drives MBAs to seek private equity jobs.

Not persuaded? OK, let’s move to religion.

What’s the mission of your local church, synagogue, mosque? Probably some form of “worship the lord and do good deeds.” Lets apply the incentives logic.

It suggests churches et al should do a religio-ethical baseline competency assessment before letting you join (“how many of the Ten Commandments would you say you break in an average month?”). Then measurements should be taken periodically to see how you are improving.

(I’ll just keep using “church” here as shorthand, please infer the intended political correctness).

If your minister/priest/rabbi shows good results, then his or her market value (salary) should increase.

Church by church results should be published, so that would-be members can make informed decisions about which church will give them the highest spiritual/ethical ROI on the least amount of invested time or payments.

If you’re a slow improver, then churches would be incented to dump you as an incorrigible recidivist (formerly known as “sinner”)—basically, an unprofitable case.

Consider tithing—the giving of 10% of one’s income to the religion. Let’s apply incentives. What’s in it for me? Maybe, if you tithe more, you get more back!  Sort of like frequent flyer miles, or volume discounts. If you tithe 12%, you get the superbowl tickets; 13%, the big Hawaii trip. And suppose you really demonstrate your holiness by taking a vow of poverty? Wow, that’s the big reward—a day’s free shopping at Neiman Marcus, no limits.

How about Boy Scouts helping that little old lady across the street? What’s in it for me? How much to escort you across, old lady? A nickel? What century you livin’ in?  Fuggedaboudit!

If you haven’t heard of Alfie Kohn, let me recommend him to you. Mainly a child educator theorist, he’s also written fascinatingly about the fallacy of incentives. As he puts it, “incentives work very well. They incent people to get more incentives.”

His key concept? Emphasizing extrinsic incentives—i.e. “if you do this, you’ll get that”—is responsible for destroying intrinsic incentives.

His killer example: a study of children playing. Researchers uncovered their favorite game—then offered them rewards for playing it.

The kids then lost interest in the game.

6 replies
  1. Ann
    Ann says:


    Your blog is always one of my "must reads" – and this is a post I feel compelled to add some thoughts to.

    Incentives are an admittedly complex field of endeavor and there is no denying that incentives are often misused and misapplied – with results that range from the merely unintended to the completely disasterous.  Alfie Kohn does us all a service by calling attention to the often misguided and flawed programs which are put in place in an attempt to improve performance.

    My issue with his perspective is that he paints all reward efforts with the same broad brush of condemnation, using arguments and examples that, in fact, represent very poorly conceived and implemented practices.  My experience would suggest that incentives also have to capability to do very good things, among them creating opportunities for employees and their employers to share in the success they create together.

    In an effort to introduce a counterbalancing viewpoint to your discussion, from what I freely concede may be a biased point of view (full disclosure – I design compensation plans, including incentives, for a living), I submit a link to a recent post of mine, Punished by (Poorly Conceived) Rewards, discussing Alfie Kohn’s book and precepts.

    Thanks for your writing!

  2. Charlie (Green)
    Charlie (Green) says:

    Ann, thank you so much for making this very timely comment.  I encourage all Trust Matters readers to click over to her own posting, where she writes specifically about this issue, providing a kind of counterpoint to Alfie Kohn’s themes–and doing so in a measured, thoughtful way.

    I don’t think the last word has been said on this issue by any means, but both Kohn and Ann have raised the level of dialog considerably.  Many thanks for "dialing in."


  3. Paul Hebert
    Paul Hebert says:

    Complete disclosure – I have spent 20 years helping companies do exactly what you stated in your opening – aligning rewards and objectives.

    I have also spent 20 years explaining the risk of unintended consequences (as you also point out.)

    I have also spent years explaining how Alfie Kohn has NO application in the business environment.  Mr. Kohn’s experiments that have been used to "show" that extrinsic rewards are bad are based on educational environments with children and activities that have an intrinisic component to them.

    The problem with transferring that wisdom to the working environment is that we are not working a.) with children and b.) on activities that create long-term benefit for the individual.

    Most incentive and reward programs when properly designed are focused on specific behaviors within a much larger context in an organization.  In other words, the incentive isn’t normally applied to a "total" activity such as in Kohn’s studies (ie: reading) but focused on a behavior that isn’t intrinsically valuable or desireable.  Does anyone find fulfillment in filling out reports or learning a new enterprise time-tracking software?  Am I really going to dampen someone’s intrinsic motivation to fill out their forms if I add a little perk to the equation?  Do I really care?

    I point you to another point of view on rewards and recognition that draws a completely contrary point of view:

    Rewards and Intrinsic Motivation: Resolving the Controversy by Cameron&Pierce. 

    While this book isn’t without its critics it provides a balance to the Kohn stuff.

    I also point you to some posts on the subject on my blog – here and here.

    I don’t disagree that for children creating incentives can create a misdirection of effort.  However, when designed properly – incentives can "align goals and incentives."

    There is also a whole discussion on the issue of using cash vs. non-cash as the reward and the difference between recognition and incentive (drives completely different behaviors.)

  4. Charlie (Green)
    Charlie (Green) says:

    Paul, thanks for throwing in some spices and mixing it up. 

    Let me push back (respectfully) a little bit.  Having spent 30 years watching consultants try to sell expensive programs to clients, I think Alfie Kohn’s ideas apply EXACTLY to business.

    When a head of systems is going to spend $10MM 5-year project, it can make or break his career.  That is non-trivial. 

    When a CEO and a head of strategic planning and a few business heads are deciding who to advise them about whether to stay in a core business or to diversify–that is serious business.   It’s as life-meaningful to them as the kid stuff is to a kid.

    And my 30 years of observation tells me that nothing sells better than a consultant who really, actually cares about the client’s success.

    And nothing sells WORSE than a salesperson who is fixated on the "win" and the commission.   The more you hype the contest and the reward, the more you kill the sale.

    To me, that’s exactly the same point as with kids.

    And yet–I see loads of companies who want to "get sales-oriented," and who mindlessly do so by believing "you can’t manage what you can’t measure."  The first thing they do is set up sales credit systems; the second thing they do is announce they’re going to "pay for performance." 

    The result is predictable: a few good people, like Huckleberry Finn, will do the right thing even when they believe it’s bad.  A few really great people know to do the right thing, even when it goes against the incentive process.

    But the mass of people come to believe the point of being sales oriented is to get the sale, and to make the money.  And it colors every interaction with the client.  And the sad irony of it is–it doesn’t work.

    Clients of companies selling complex and intangible goods don’t want people who are running self-motivated calculators in their heads; they want people who give a damn and are willing to look at the long haul and over multiple events.  The way incentive comp gets played out, that’s not what happens.  It’s the deal, the transaction, the quarterly numbers, the bonus to me, me me.

    While I think a misalignment between reward to the seller and reward to the buyer is long-term unsustainable, I’d also say that on the whole and on the average in business today, we see an excess not of intelligent alignment, but an excess of monetary-based short-term alignment as fix.

    I’m not disagreeing in principle, but in fact.  Empirically speaking, the practice of linking money to short-term outcomes is overused, over-applied, and over-credited.

    Or so it seems to me.


  5. Paul Hebert
    Paul Hebert says:

    I totally get your point of view and don’t worry about "respectfully" pushing back – I love the discussion.

    Your example is exactly what I tell clients not to do – DO NOT use incentives to reward sales if your real issue is building trust and making recommendations with integrity.  In your example rewarding the sale is wrong.

    By focusing on the behaviors and rewarding the steps to the sale such as demonstrating value, exploring options, providing case histories, offering references, etc. will drive the appropriate behaviors.

    I’ve posted before on the issue of rewarding results not behaviors.  That is where the problem typically begins.

    I also am in vehement agreement on the use of incentives as a stop-gap fix.  I liken it to the overuse of antibiotics in today’s fight on infections.  We have the same problem in business – the rampant application of poorly designed incentives has poisoned the entire business world.

    Properly applied incentives make sense.  Poorly applied they don’t.


Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *