Carrots and sticks
The mismatch between what science knows and what business does
There is a time to admire the grace and persuasive power of an influential idea, and there is a time to fear its hold over us. The time to worry is when the idea is so widely shared that we no longer even notice it, when it is deeply rooted that it feels to us like plain common sense. At the point when objections are not answered anymore because they are no longer even raised, we are not in control: we do not have the idea; it has us.
Punished by rewards
These were the opening words from Alfie Kohn’s book ‘Punished by Rewards – The trouble with gold stars, incentive plans, A’s praise, and other bribes’. These words rang a bell with something I had been questioning for some time. My studying of, and work with, Marshall Rosenberg’s concept of Non-Violent Communication had lead me to question the use of superficial praise such as ‘thanks for the great presentation’ or ‘that’s an excellent report’.
When I discuss the shortcomings of such communication with people in my close environment, they can relate to what I say, but nonetheless appear content, or even get a short-lived kick, out of such superficial praise. I can’t help but think that we are missing out on so many opportunities to have more meaningful and life-enriching communication.
So it was with much interest that I picked up Alfie Kohn’s book, in my search of a better understanding of the impact of rewards and praise in our society and in particular in our workplaces. In a book packed with notes and references to articles and field studies I found what I was looking for.
Around the time that I was half way through Kohn’s book, I was out running one evening with my friend Matt. I shared some of Kohn’s thinking with him. He pointed me towards a TEDTalk from Dan Pink entitled The surprising science of motivation. Pink’s talk lead me in turn to pick up a copy of his book Drive.
This blog synthesises what I took from these two books. In writing this blog post I have extensively used extracts from the original texts.
Praise – the dominant mindset
The dominant mindset that has established itself over the past two decades is that we should refrain from punishing and criticising and instead attempt to “catch people doing something right” and reward them with privileges or praise. This mindset is popularised in the management books and seminars of the day. The underlying assumption is that there are exactly two alternatives: positive reinforcement or punitive response, otherwise referred to as carrots or sticks.
When choice is framed this way only a simpleton would fail to pick the carrot. Rewards are less destructive than punishments, and the difference between the two becomes more important as the punishment in question becomes more harsh. But the dichotomy is a false one: our practical choices are not limited to two versions of behaviour control. And that is very good news indeed because despite the relative superiority of rewards, the differences between the two are overshadowed by what they share.
Although many people counter-pose rewards to punishments, it is interesting to observe that the two strategies often go hand in hand in the real world. In a study reported in 1991, elementary school teachers from thirteen schools were observed over a period of four months. It turned out that the use of rewards and punishments in the classroom were very highly correlated; the teachers who used one were more, not less, likely to use the other.
Other studies have found that even praise, the form of reward usually viewed as the least objectionable, is often favoured by people whose style of dealing with children is conspicuously controlling or autocratic.
How praise drags down achievement
Kohn provides four reasons why praise not only fails to boost achievement but actually drags it down:
- Signals low ability: If the person being praised considers the task easy, then he interprets the praise to mean that he isn’t smart. This inference triggers low expectations and in turn low persistence and performance intensity.
- Makes people feel pressured: Telling someone how good she is can increase the pressure she feels to live up to the compliment. This pressure, in turn, can make her more self conscious, a state that often interferes with performance.
- Invites a low risk strategy to avoid failure: Studies have shown that there is a high correlation of the use of rewards and punishment. In the same vein those that heap praise may also be generous with the dishing out of criticism. There is even a method known as the sandwich technique that encourages people to wrap a compliment around (i.e. before and after) a criticism.
- Reduces interest in the task itself: The attention of the person being praised is deflected away from their intrinsic satisfaction toward the extrinsic praise.
Rewards come in many shapes and sizes, with praise being only one form. With the help of a number of field studies I have highlighted below the impact that other types of rewards (and punishments) have on motivation.
The impact of rewards on learning
Anything presented as a prerequisite for something else – that is, as a means towards some other end – comes to be seen as less desirable. The recipient of the reward figures: “If they have to bribe me to do this, it must be something I wouldn’t want to do.” Promising a reward for an activity is tantamount to declaring that the activity is not worth doing for its own sake. And the greater the incentive we are offered, the more negatively we will tend to view the activity for which we receive it.
Field study – Killing curiosity and our innate desire to learn new things
Borrowing a technique from Edward Deci, Oprah Winfrey staged an experiment in 1996 and asked Alfie Kohn to come onto her show to explain the results:
Twenty children were invited into an office building one at a time and were greeted by someone who pretended she worked for a toy company. Each child was invited to help evaluate some new puzzles. Half the kids were promised a reward of five dollars for each puzzle they tested. After the playing and evaluating were done, each child was left alone in the room for a few minutes – and secretly video-taped. It turned out that every one of the ten kids who had participated in the evaluation without any mention of payment, went back to playing with one of the puzzles when the formal testing period was over and no one was around. But of the ten kids who had been rewarded for participating, nine did not touch a puzzle again.
The impact of rewards on altruistic behaviour
The use of rewards also has an impact on the extent to which people exhibit altruistic behaviour. The following three field experiments illustrate the point nicely:
Field study – Comon good
Frey and Oberholzer-Gee (1997) surveyed citizens in Swiss cantons where the government was considering locating a nuclear waste repository.
They found that the percentage of people supporting locating the facility in their community fell by half when public compensation was offered.
Field study – Donating blood
In a similar line of logic a study in Sweden by Mellström and Johannesson showed that after the introduction of a financial incentive to donate blood, the percentage of those willing to donate dropped from 43% to 33%.
If the option of donating the payment to charity is built in then the willingness to donate goes up to 44%.
Field study – Concientious parents
In 2000, the economists Uri Gneezy and Aldo Rustichini ran an experiment were they announced the imposition of a fine every time parents were late picking up their children from a childcare facility.
After the introduction of the fine they observed a increase in the number of the parents coming late, settling at a rate twice as high as before the introduction of the fine.
The fine had deflected the parents attention from the intrinsic desire to be considerate to the childminders, to a rational transactional decision – is the value of the extra personal time greater than the cost of the fine?
The impact of rewards on creativity and technical skills
“Do this and you’ll get that” makes people focus on the “that” not the “this”. Prompting employees to think about what kind of bonus they could earn this year is about the last strategy we ought to use if we care about creativity. Said another way: Do rewards motivate people? Absolutely. They motivate people to get rewards.
Rewards usually improve performance only for simple tasks, and even then they improve only quantitative performance.
The trouble is that we have incorrectly inferred from this fact a general law of human nature – reward people and they’ll do a better job – and applied it in our workplaces and schools.
Field study – How to cloud thinking and dull creativity
In 1930 the psychologist Karl Duncker devised the “candle problem”.
The participants are presented with the materials as seen in the illustration and set the objective to attach the candle to the wall so that the wax does not drip onto the table.
After 5-10 minutes most participants work out the solution as depicted in the second illustration below – namely placing the candle in the emptied box which has been fixed to the wall using the thumbtacks. To achieve this result participants needed to overcome what is called “functional fixedness”, meaning they need to see beyond the thumb tack box as purely a container for the thumbtacks.
He divided the participants randomly into two groups. The first group was informed that they would be timed to establish norms for how long it would typically take people to solve this sort of puzzle. The second group of participants were each offered $5 if the time they took to solve the problem was in the top 25% of all those tested. If it was the fastest the reward would be $20.
The results, which may be a surprise to some, but probably not to those interested in reading a blog on this subject, were that the extrinsically incentivised group took an average of three and a half minutes longer. This happens because rewards narrow people’s focus, blinkering the wider view which is necessary to solve less straight-forward problems.
But then Glucksberg took Duncker’s experiment one step further. He presented the problem with the thumbtacks emptied onto to the table next to the container.
This time the results were reversed. The participants who had been offered the reward solved the problem faster.
By leaving the thumbtack box lying empty on the table, Glucksberg had changed the puzzle from a heuristic task (i.e. a task that required experimenting with possibilities to devise a novel solution) to one which was somewhat algorithmic in nature (i.e. more akin to following a set of simplistic steps down a single pathway to one conclusion).
To summarise: incentives will have a detrimental effect on performance when two conditions are met:
- the task is interesting enough for subjects that the offer of incentives is a superfluous source of motivation;
- the solution to the task is open-ended enough that the steps leading to a solution are not immediately obvious.
Incentive and merit pay plans
The impact of competitive rewards
Kohn takes the view that of all the ways by which people are led to seek rewards, one of the most destructive arrangements is to limit the number that are available. Placing a constraint on the number of rewards (e.g. bell-curve style salary increases) can cause the following effects:
- destructive impact on teamwork
- competitive anxiety that can interfere with performance
- those that don’t believe that they have a chance of winning are discouraged from making an effort
Why incentive and merit pay plans fail
- Lack of necessity: Plans are sometimes introduced when employees are already doing a fine job. And if the task is interesting enough for subjects then the offer of incentives is a superfluous source of motivation;
- Secrecy: When no one is supposed to know how much anyone else makes, people may overestimate what others actually earn or assume that inequities exist even when they don’t;
- Pay doesn’t match performance: For a variety of organisational reasons or by design of the merit plan itself, compensation often does not correspond to performance ratings;
- Expense: These plans cost money, money which could be invested to better meet the needs of the organisation and employees;
- Too big versus too small: If bonus payments are modest, they may not have much of an impact. If they are large enough to make an impact, fewer people will receive them;
- Short-term versus long-term: If an incentive is based on short-term performance, employees may make decisions counter to the long-term interest of the organisation. If the incentive is based on performance over a longer period of time, the connection between behaviour and reinforcement becomes fuzzy;
- Objective versus subjective: If compensation is based on objectives factors, the system may be rigid and unresponsive to those aspects of performance that defy reduction to a fixed protocol. If compensation is based on subjective factors, it becomes dependent on the whims and biases of the rater;
- Difficulty in assessing meaningful graduations in quality: It is difficult to quantify performance for many kinds of work and most rating systems are accurate only at the extremes – that is for identifying exceptionally good and bad performances;
- Pay is not a motivator: Several studies have found that when people are asked to guess what matters to their coworkers – or in the case of managers to their subordinates – they assume that money is at the top of the list. But put the question directly – “What do you care about?” – and the result looks different;
- Rewards punish: Rewards feel punitive because they too amount to an effort to manipulate people’s behaviour. Additionally, employees perceive rewards that they failed to obtain as punishment. Withheld rewards can create an even stronger perception;
- Rewards rupture relationships: Incentive plans that encourage peers to compete for scarce rewards, damage teamwork. Having managers decide on employee bonuses encourages employees to hide problems and avoid challenging poor decisions;
- Rewards ignore reasons: Applying rewards rather than understanding what is really happening is easier, albeit less effective (e.g. why are the two teams are duplicating work; why does the presentation of results looks impressive but the back office is a mess; why senior management were not aware of the creative accounting that has been in place for over a year; why is there a recent rise in sick leave);
- Rewards discourage non-hopefuls: Those that don’t believe that they have a chance at winning are discouraged from making an effort;
- Rewards discourage risk-taking: If people get rewards for completing tasks, they choose the easiest possible tasks. In the USA an incentive program that rewards children with pizza vouchers for every book they read has lead to shorter and easier books being red;
- Rewards undermine interest: By moving the focus away from the activity and towards the reward, the intrinsic motivation gets diluted.
Minimising the damage of incentive and merit pay plans
If you cannot get rid of rewards Kohn provides the following recommendations to minimise their damage:
- Get the rewards out of people’s faces;
- Offer rewards after the fact, as a surprise (rather than ‘if then’, use ‘now that’ rewards);
- Never turn the quest for rewards into a contest;
- Make the rewards as similar as possible to the task (if you feel compelled to give a child something for having read a book, give her another book);
- Give people as much choice as possible about how the rewards are used;
- Do what you can to help people shrug off the implicit message offered by extrinsic motivators.
In other words, pay people generously and equitably. Do your best to make sure they don’t feel exploited. Then do everything to help them put money out of their mind.
Changing the dominant mindset
The use of punishments and rewards are appealing because they do not require any attention to the reasons why the subject is not conforming to your wishes. You don’t have to ask why the employee is indifferent to her job, produces reports littered with typos, shows little interest in learning or doesn’t know the meaning of the word initiative.
Kohn illustrates this well with an example from child rearing. One wonders how anyone could presume to propose a solution without knowing why a child keeps getting up after having been put to bed. With very little effort we can imagine several reasons for this behaviour. Maybe she’s been put to bed too early and simple isn’t sleepy yet. Maybe she feels deprived of quiet time with her parents, and the evening offers the best opportunity for her to cuddle or talk to them. Maybe she’s still wound up from what happened a few hours earlier and needs to rehearse and clarify the day’s events to try to make sense of what happened. Maybe there are monsters under her bed. Or maybe she can just hear people talking in the living room.
The point is that we don’t yet know what’s really going on. But the punishment and reward based approach doesn’t require us to know. Each of the possible explanations for why this girl doesn’t stay in bed at night would seem to call for a different solution. This is one reason it is difficult to give a simple answer to people who demand to know what ‘the alternative’ is to using rewards. Rewards are not actually solutions at all. They are gimmicks, shortcuts and quick fixes that mask problems and ignore reasons. They never look below the surface.
The impact of getting it right
If we pause for a brief moment, as we come to the end of this blog, to understand the impact of increased motivation, we can see that it will influence learning, altruism, ethics, creativity, well-being and productivity to name but a few dimensions. But how much impact are we talking about? Is it 10% more creativity, 20% more learning and 30% more productivity?
Without going into any studies on this, have a look for yourself. Think of the following two phases or situations in your life to date. The first situation is a time when you were in a peak state, loving what you were doing, creating results like you have never done before, building your capacity to do more, and experiencing the moment with enthusiasm. Then think of time when your were disengaged, dragging you heals, failing to see the purpose, bored with the activity and not personally interested the the result. What was the difference in terms of learning, creativity and output? I know for myself that the difference is not 10% or 30%. The difference is like night and day, and is certainly a multiple rather than a percentage of a difference.
It is possible to get people to do things. That is what rewards, punishments and other instruments are all about. But the desire to do something, much less do it well, simply cannot be imposed. In this sense it is a mistake to talk about motivating other people. All we can do is to set up certain conditions that will maximize the probability of their developing an interest in what they are doing and remove the conditions that function as constraints. And to increase the probability that we get these conditions right, we need to look below the surface.
To learn more
Both Kohn and Pink offer alternative approaches to the carrots and sticks model. With these and other approaches we go below the surface and broaden our perspectives and understanding of the fundamental human needs that drive our each and every action. This topic is worthy of a dedicated blog post;). In the meantime if you would like to discover the models that Kohn and Pink present, I can recommend their books. ‘Punished by Rewards’ is a comprehensive read which focuses equally on parenting, teaching and the workplace. Dan Pink’s book is a lighter, shorter and more workplace orientated read.