Book Review/Notes: Punished By Rewards
Here are some thoughts and highlighted sections from Alfie Kohn's book Punished by Rewards: The Trouble with Gold Stars, Incentive Plans, A’s, Praise, and Other Bribes. It's an interesting book that will make you think about how you interact with people. This article is more "quickly sharing some interesting ideas" and not "highly polished essay". Sections in italics are quotes from the book.
A few times lately I've come across a quote about "respect". It goes something like this: sometimes people use the verb 'respect' to mean 'treat someone like an authority'. Other times it means 'treat someone like a person'. So when someone says 'if you don't respect me, I won't respect you', what they really mean is: if you don't treat me like an authority, I won't treat you like a person.
Punished By Rewards has a similar energy:
A great many adults who complain that children don't act "responsibly" really mean that they don't do exactly what they're told.
Rewards imply the thing isn't worth doing
Kohn has a few reasons why we should be anti-rewards. The most memorable reason: if you're rewarded for doing something, that's a subtle message that the thing is not worth doing on its own. Pay a kid $1 for each book to read, and they'll start to think that books are something you read for money, not because reading is fun. In scientific terms: extrinsic rewards reduce intrinsic motivation.
If you feel rewards are really necessary, see if you can "reduce the gap between what people are doing and what they are getting for it. If you feel compelled to give a child something for having read a book, give her another book." The other option is giving rewards that are unexpected: "most studies have found that unexpected rewards are much less destructive than the rewards people are told about beforehand and are deliberately trying to obtain. "
On a similar track, when you offer a reward for good behavior, it can be seen as a subtle hint that you think the person isn't capable of good behavior without the reward, which is kind of insulting:
If the capacity for responsible action, the natural love of learning, and the desire to do good work are already part of who we are, then the tacit assumption to the contrary can fairly be described as dehumanizing.
Other reasons to be against rewards
1. Rewards are essentially punishments — "do this and you'll get a reward" is very similar to "don't do this and you'll get punished". In either case, you're motivated to do this because your life will be worse if you don't. In both cases you're trying to avoid the hypothetical future where your life is relatively worse and end up in the hypothetical future where your life is relatively better. It just a difference of what we're considering the "baseline" in that particular case.
2. Rewarding some people but not others fosters competition in situations where you might prefer collaboration. On top of that, if some people don't think they'll be able to win the reward over their more-skilled peers, they might be discouraged from even trying.
3. Rewards disrupt the relationship between the giver and the receiver. Instead of having an honest, equal relationship, the potential reward receiver will play the game of trying to get the rewards. An employee has more motivation to hide an issue from their boss. A punished child can't help but partly see their parent as the source of punishments.
A parent or teacher who relies on punishment becomes, in the eyes of the child, a rule enforcer, someone who may cause unpleasant things to happen—in short, someone to be avoided.
4. Rewards ignore the deeper level. Most if not all rewards have to be based on some explicit goal. But that goal doesn't necessarily capture the deeper goals we have in life (similar to Goodhart's Law). Rewarding someone for avoiding a certain behavior ignores why they were tempted to engage in the unwanted behavior in the first place. Rewarding someone for grades ignores whether they learned the material in a lasting way.
Children are encouraged to focus in a legalistic way on exactly what behavior is covered by each rule, how the rule will be applied, what circumstances may create exceptions, and so forth.
Sometimes we do have to do boring things
So how do you get an employee/child/etc to do something boring if not for rewards? Here's the suggestion:
First, imagine the way things look to the person doing the work and acknowledge candidly that it may not seem especially interesting. Second, offer a meaningful rationale for doing it anyway, pointing, perhaps, to the long-term benefits it offers or the way it contributes to some larger goal. Third, give the individual as much control as possible over how the work gets done.
It's all messed up
If the only way people are willing to do the current authority's bidding is bribes or threats, that authority must be asking them to do something that's not very fulfilling. The systems of our society (how schools are run, how companies are run) are incompatible with people being intrinsically motivated. A few sample quotes:
"Idleness, indifference and irresponsibility are healthy responses to absurd work," said Herzberg. Elsewhere he issued this challenge to executives: "If you want people motivated to do a good job, give them a good job to do."
One of my epiphanies as a teacher came with the realization that students' disruptive acts were less a sign of malice than of a simple desire to make the time pass faster. No strategy for classroom management can hope to be effective in the long term if it ignores the fact that misbehavior often reflects students' lack of interest in much of what we are teaching. They can't get out, so they act out.
For whom do rewards work best? For those who are "alienated from their work"... If what you've been asked to do seems silly or simple, you might decide to make a real effort only when there is something else, something outside the task itself, to be gained.
the fact that [reward-and-punishment] strategies are invariably used to promote order and obedience would have to be explained as a remarkable coincidence. Rewarding people for making changes in the existing order (which might include the very order that allows some individuals to be controllers and others controlled) is not merely unlikely but a contradiction in terms.
Some managers justify a focus on high wages by saying that this is what employees ask for, but:
There is a tendency to focus by default on the size of one's paycheck when work is bereft of more important features: deprive someone of a genuinely engaging and meaningful task, the capacity to exercise choice over what one does, social support, the chance to learn and to demonstrate one's competence, and that person will likely turn his attention to what he earns. (He may even dismiss as naive the suggestion that work could ever be about more than money.)
Sometimes people impose control because they're themselves stuck in a messed-up system:
So why don't teachers stop doing these things? I think there are several plausible explanations. First, some teachers cannot imagine how else they could do their jobs, particularly when they must work with children whose behavior is difficult to deal with—or for that matter, when they simply have too many children in one room. Controlling academic strategies, in other words, can be a response to nonacademic features of the classroom.
Second, it takes more time to bring students in on the process of making decisions, and many teachers already feel there are not enough hours to do what has to be done.
Having Control
This quote came at the end of the book, but in some ways it underlies it all:
Many different fields of research have converged on the finding that it is desirable for people to experience a sense of control over their lives.
For example, here's a parenting example:
In the case of what is usually referred to as a disciplinary matter, the first stage [in a truly collaborative approach] is to talk about whether the act in question really is a problem, and if so, why. Typically parents make this determination on their own and announce that the child's behavior must change. The child, however, may not understand why it should change—other than because a powerful person, with rewards and punishments at his command, decrees it...Is it really wrong for children to keep their rooms messy?
Or:
Why do people so often mutter and groan when a change in policy is announced? Not because it is a change but because it is announced—that is, imposed on them. (This distinction forces us to reconsider assumptions about the alleged conservatism of "human nature." As Peter Scholtes has observed, "People don't resist change; they resist being changed.")
The important twist here is that even if you're convinced by the book that rewards are bad...you (as a manager/parent/teacher) shouldn't change things without getting buy-in from the people you're in charge of.
On a similar level, I think people respect us when we communicate with them in fundamentally honest ways. When we ask people sarcastic/hypothetical questions that they know we're not interested in actually hearing an answer to, other people know what we're doing. They resent us for taking away their control and choice.
Sometimes the question is posed, but in a harsh, threatening tone ("Why are you late getting ready for school again? Huh? Why? Answer me!"), so the child understands full well that he is not really being asked to speculate about possible reasons; he is being asked to drop his head, look penitent, mumble something, and hope to avoid punishment.
Avoiding these conversational patterns that take away someone's control is a key to being nice, in my opinion. If you're not actually curious about the answer, don't ask the question.
Praise is a type of reward
Praise is not exempted from the anti-reward argument.
Most books on parenting offer no hint that positive comments could be anything but constructive...
but...There are a few anti-praise arguments. First, praise often implies an implicit difference in status between the praiser and the praisee:
the most notable aspect of a positive judgment is not that it is positive but that it is a judgment..."It is interesting to note here that when the work of a high-status person is praised by a low-status person, this is often seen as presumptuous or even insulting," one writer points out. Precisely because praise usually implies a difference in position, it can produce resistance.
...in the typical ceremony for "recognizing excellence," the people in charge have unilaterally selected, at their own discretion and based on their own criteria, some people to recognize over, and in front of, others. It is their power to do so that is ultimately being recognized.
And public praise can lead to awkward dynamics between those who receive praise and those who don't:
Competition is also fostered by giving praise publicly. For example, the elementary school teacher who announces in front of the class "I like the way Stewart is sitting so nice and quiet and ready to work" has set up a contest for Nicest, Quietest Student, and everyone other than Stewart has just lost...the interaction is fundamentally fraudulent because the teacher, while pretending to address Stewart, is actually using Stewart to manipulate the behavior of others in the room.
Perhaps more importantly, praise can lead to being unwilling to take risks:
One classic classroom study found that students whose teachers frequently used praise showed less task persistence than their peers. Why? Perhaps because praise sets up unrealistic expectations of continued success, which leads people to avoid difficult tasks in order not to risk the possibility of failure.
Aside from not praising publicly, Kohn offers these suggestions:
1. Don't praise people, only what people do...
2. Make praise as specific as possible... (And it doesn't even have to be praise, per se:
Just pointing out an aspect of a child's essay or drawing that seems interesting (without saying that it's nice or that you liked it) will likely be sufficient to encourage her efforts.)
3. Avoid phony praise... Praise becomes objectionable when it is clearly...a deliberate strategy
Overall: With every comment we make—and specifically, every compliment we give—we need to ask whether we are helping that individual to feel a sense of control over his life.
Can we successfully determine who deserves a reward?
Another reason to doubt the effectiveness of rewards is that we should question whether we can even successfully determine who deserves it. One employee taking credit for another employee's idea is a common enough idea that it has a TV Tropes page.
In today's world, many of the things we work on are extremely complicated and rely on large numbers of people working together. Can we accurately separate out one person's contribution from another? As Kohn puts it,
It is nearly impossible to quantify performance for many kinds of work, and in any case, most rating systems are accurate only at the extremes—that is, for identifying exceptionally good or bad performance. That means gradations in compensation will be made that probably don't correspond to meaningful gradations in quality.
I have mixed feelings about this. I see some truth in it. But I've also been in situations where I felt like I did more than my fair share of the work—and it sort of feels like a "punishment" to be given the same reward as everyone else when you've done a better job than them. I'm not sure there are any easy answers to this question. The best I can say for now is that the world is complicated and we won't always get exactly what we deserve (plus we're all biased—we won't always deserve exactly what we think we deserve). It's best to accept that and understand it'll happen sometimes, while still being willing to stick up for yourself in the face of consistent bias or gross unfairness.
He also points out that often our highest performers are people who are passionate about what they do. Being in it for the money is often seen as a lesser calling than to do something out of passion. Shouldn't we encourage that truth by avoiding reward systems that encourage people to care about the money:
individuals who are committed to excellence and likely to do the best work are particularly unlikely to respond to financial incentives.
So what do we do instead?
Although he does spend some time discussing what a world without punishment or rewards looks like (giving people choices; having honest conversations; providing feedback; building better, more innately satisfying systems), I'm not sure it's enough.
Punished by Rewards could benefit from having a companion book of parenting advice that actually covers specific situations (perhaps that book already exists by now). It's fine to say that our goal should be having actual conversations and getting buy-in from the people involved. But there's so many specific situations where it's hard to think of the non-rewarding, non-punishing way to fix those issues. And there can be a balance involved, too, especially where younger children are involved:
This raises the idea of control, a word that generally implies the use of coercion or pressure to impose one's will on a child. Edward Deci and Richard Ryan have distinguished an environment for children that is structured or limited from one that is controlling. Their objection is to the latter. Even the most ardent opponent of control would hesitate to rule out all such interventions, particularly with young children. Sometimes it is necessary to insist, to put one's foot down. Still, it seems reasonable to propose that parents and teachers try to use the least intrusive or coercive strategy necessary to achieve a reasonable end. Don't move a child roughly if you can move her gently; don't move her gently if you can tell her to move; don't tell her if you can ask her.
I'm not completely clear on what would distinguish a "structured" environment from a "controlling" environment in every case. I respect that he admits that a little control might be necessary, especially with younger children—that matches my personal intuition based on how crazy some of the younger kids I know can get. I guess the key is using the minimum possible control and transitioning into a system of honest conversations and buy-in as children grow up instead of continuing to use control.
(Coincidentally, I read this book right after reading a Cesar Millan book. Dog trainers have some shared philosophies with this book—to paraphrase Cesar, you need to recognize that your dog is a dog and has doggy needs and not human needs. But overall, positive reinforcement works much better for dogs because they're not on the mental level of humans.)
Another part of the answer to the question of "what do we do": we can provide pure feedback, instead of feedback attached to a reward or punishment:
...informational feedback is an important part of the educational process. But...reducing someone's work to a letter or number is unnecessary and not terribly helpful. A B+ at the top of a paper tells a student nothing about what was impressive about the paper or how it could be improved. A substantive comment that does offer such information, meanwhile, gains nothing from the addition of the B+.
Overall, though it's a little repetitive (explaining how rewards are bad, then explaining how rewards are bad in the classroom, then explaining how rewards are bad in the office…), the book has a lot of interesting ideas. I'll definitely keep these concepts in mind in my future as a coach, parent, or manager.