Why rewarding students with money is a bad idea.
And what it means for teachers.

There’s a story told by a writer, trying working from home, who was so distracted by the noise from kids playing joyfully in the park by his house that he couldn’t concentrate. As a keen student of human nature, he decided to tell them that he loved their noise, and he would pay them each a dollar a day to keep it up. After a week, he told them he was reducing it to 50c; and shortly afterwards, he said he could no longer afford it, and they were free to keep going if they wished, but it would be for free. The kids drifted away, and the writer wrote in peace.

So what’s the best way of stopping children having fun? Pay them to do it, and then stop paying them. Make it a task to be completed to earn a reward; take away the joy.

There are lessons here for schools, which run on a similar combination of social psychology and economic theory. The equivalent question is What’s the best way to stop students learning? The answer is not so different.

We want to motivate students and teachers to give of their best; and like any organisation, certain things must and must not happen for thousands of people, machines, pieces of equipment and processes to mesh together in an effective way. That infants and teenagers are involved makes it even harder (as well as more fun). 

So how do we go about this business of promoting the behaviours we want and need, from adults and children? 

This is a far more important question than it might appear – because often it is never asked, with the answers simply assumed as obvious; and also because one’s fundamental orientation here will in large part determine how students and teachers experience schooling. It will shape the identity of the organisation.

The most obvious answer about is the carrots-and-sticks one; reward people for doing well, and punish them for doing badly. However, it turns out that it’s one of those obvious answers that is actually wrong. In fact, there is so much wrong with it as a basic approach that it is hard to know where to begin. 

Morally,  the carrots-and-sticks approach seems to treat people as objects to be manipulated to perform certain action in certain ways; it does not consider why people do things, or how they feel about what they do, or address their long-term development or interests (which is doubly terrible in a school, of all places). Based in the now discredited behaviourism movement, it feels rather of the ‘children should be seen and not heard’ line of thinking which sits so poorly with current progressive notions of parenting and schooling.

But even leaving the moral argument aside, the evidence is that the carrots-and-sticks approach simply does not work to generate the behaviours we seek. As far as the stick part goes, the evidence in favour of non-punitive restorative justice over traditional punitive approaches is increasingly robust (see here and here for example) – but it’s on the carrots that I want to focus here.

Carrots – or rewards, incentives, call them what you will – are incredibly seductive. How could rewarding people for doing the right things be wrong? It seems so natural – indeed almost axiomatic to many. It even seems to be enshrined in the most fundamental economic ‘law’ – that raising (monetary incentives) increases supply and production. But the ‘laws’ of economics are far more complex than the ‘laws’ of physics, and far less universal. As a result, the narrow economic approach simply misses the point. In particular, it simply misses the subtlety of how choice works.

We want students and teachers to freely choose certain actions, and in doing so, choose to develop certain habits. In particular we want them to develop the habit of choosing the actions they think are right, even if they are difficult, and even when no one is looking; that is, we want the choices to be genuine ones. It turns out that the very act of reward can inhibit this longer-term habit. The classic example comes from Titmuss (2000) who describes how many blood donors stop giving blood after the introduction of rewards for donations. It’s worth stopping and thinking here – because it says something fundamental. If you wanted to give blood for free, wouldn’t a bit of cash make it even more attractive?

It turns out not. It turns out that when otherwise wholly volitional behaviours come to be seen as something done for external reward, the original reason for doing them (in this case the intrinsic satisfaction of helping others) can be ‘crowded out’. Frey (2001) describes this same phenomenon in many disparate areas where carrots are used to incentivise behaviour, but have the perverse effect of decreasing it. From patients’ readiness to take prescribed medication to the reaction of managers to various forms of supervision by their superiors; to the readiness to offer voluntary work; even the readiness to accept nuclear waste repositories, the evidence is consistent. Most importantly for us in schools, Frey marshals empirical evidence that backs up what others (eg Kohn 1993) have long said; that childrens’ learning behaviour is hindered, not supported by rewards.

So what underlies this counter-intuitive finding? One theory is that carrots signal that an activity cannot be rewarding or interesting in its own right and therefore undermines the perception of the value of that activity. In fact, it seems that carrots are essentially not rewards but bribes. No one would ever bribe anyone for something they want to do anyway, right? And if you stopped me, say, when I was going for a run, and offered me $50 to complete it, I would be naturally suspicious that there was something wrong, and that you knew something that lay ahead of me on my run, something that I was unaware of. Something bad. And so while I might complete the run to get the $50, I’d be unlikely to enjoy it so much. I might not even complete the run, despite the carrot. Even if I did, I might not run that way again, or indeed I might come to see running differently.

For students, then, rewards for performance at school would say ‘hey, this is worthless on itself but I can make it worth your while’. For teachers, rewards for performance seem to miss the professional aspect that any decent teacher is already motivated to support students. In both cases, rewards would be utterly uninspiring; at best they would generate short-term compliance to meet the carrots; at worst they would undermine the very attitudes and habits we want. Compliance is seductive because compliant behaviours look like success, but compliance is dangerous because it is at odds with developing independent individuals who are in charge of their own learning.

I am referring to schools here, and the view that neither carrots nor sticks do what we want them to do. That said, perhaps there are places where carrots-and-sticks may work. I am thinking of my time as a teenager stacking shelves in supermarkets. I was not motivated by anything other than the money; the work had no intrinsic motivation for me. I turned up, did the hours, took no real pride in the work and while I did it honestly, I did it to the absolute minimum requirements. Perhaps it would have made sense to pay me by the number of shelves I could get done, rather than by the hour (even there I have doubts; I suspect that offering me choice and the chance to work in a team would have been more motivating for me). So when the work is intrinsically unattractive and dull, and if we cannot find ways to make it interesting, perhaps we need carrots; and perhaps that’s the right model for motivation.

But this model for motivation gets carried far beyond it’s proper domain, to places where it really should not belong, because the work is not simple, and not short-term. Many conversations in schools – about ‘behaviours’, ‘assessment’, ‘discipline’, ‘awards’ and ‘performance-related-pay’, for example – are about this issue, sometimes buried deep in layers of assumptions.  Perhaps the best single-line summary comes from Ryan and Deci’s authoritative overview: rewards may lead to improved performance on simple tasks but impair performance on more complex tasks.

It’s dangerous to extend a model of thinking beyond its useful domain. This is one of those times, and I am reminded of Maslow’s Law of the Instrument: When all you have is hammer, everything looks like a nail. It may sound harmless if we disguise our hammer as carrots and sticks – but it’s still a hammer, and people are not nails.