top of page

Can money actually be detrimental to performance? 
The Candle Problem

This question sounds pretty ridiculous. Money is universally desired. And for many people, money is a big consideration for the work they do. So how is it possible that paying someone extra money could lead to a drop in performance?

There's also a deeper implication: we often use money as the carrot to encourage better performance at work. If money is detrimental to performance does this mean that big bonuses actually have a reverse effect, causing performance to drop?

Let's start with a light experiment. In 1945, Karl Duncker came up with the Candle Problem.

1_0RcyiHkLSA2Z1B94hc0XyA.png

The Candle Problem

You are given:

1 Candle

Some matches

Some thumbtacks
 

The goal is to attach a lit candle to the wall, in a way where the wax does not drip down onto the table or floor below.

How would you do it? (if the formatting is pissing you off, I admit it. I'm trying to make this paragraph as long as possible so you can't see the solution until you scroll down)

 

It turns out that everyone has some ideas. Some would try to thumbtack the candle onto the wall. Others tried to melt the candle wax, and use the melted wax as a glue. None of this works, but eventually, after some time, most people figure it out. This is the solution:
 

candle-problem-e1499721185585.png

The Solution

So this is the one solution that works. You use the box containing the thumbtacks as a base to hold the candle. And then you use the thumbtacks to attach the box to the wall. That's it.

The trick is to overcome what scientists term "functional fixedness", a cognitive bias where people are fixed in how they think an object can be used based on how it was first presented to you. For example, when the tacks were inside the box, it's easy to think of the box can just a container. In fact, some people might not even notice it. In contrast, when the items were presented as they are below - the only difference being that the thumbtacks were placed outside of the box, people got the correct answer much quicker. 

contentItem-1002796-5686117-sq94xz97710c

Another way to think about this is that it requires some creativity to overcome functional fixedness, to see other uses for the same object. Even though the solution is exactly the same, the person who solves the Candle Problem went the thumbtacks were in the box has shown more creativity than a person where the thumbtacks were already out of the box. 

Ok. At this point, you might be thinking, what about the money? No one's interested in candles, ok? Money is much more fun to talk about. If you're not talking about the money, I'm going to stop reading. 

Relax. We're gotten here. Princeton psychologist Sam Glucksberg read about the Candle Problem and wanted to test out if money would incentivise people to solve the problem faster. 

addq.png

Glucksberg divided participants into 2 groups:
 

Group A: was just asked to solve the Candle Problem - no incentives were offered even if they solved it.

Group B: participants were told that if they were among the fastest 25% to solve it, they would get a monetary reward. The fastest to solve the problem would get an even bigger monetary reward. 

Now if money is a motivator as we suspect it to be, then we would expect Group B to finish the task faster. 

dasdsa.png

But the results are completely opposite. Group B, the group which was offered a monetary reward performed significantly slower than Group A. 

This has some important implications, but hold on, we'll get back to this. There's one variation to share.

In the experiment above, the participants solved the actual Candle Problem - the more difficult version where the thumbtacks were in the box, and participants are more likely to suffer from functional fixedness. What if participants did the easier version of the Candle Problem - where the box was presented separately, removing functional fixedness?

dwqdwqwd.png
dwdqdw.png

Now we see that the results are reversed. 

In the simpler version of the Candle Problem (with no functional fixedness) - the group offered a monetary reward performed better. But in the more complex, original version of the Candle Problem (with functional fixedness), the group offered the monetary reward performed worse.
 

What's happening here? 

It turns out that the monetary reward naturally causes us to narrow our focus - let's quickly finish this so that we can get the money. This might work for routine issues, but for more complex problems that require creativity (i.e. the original Candle Problem), the excessive focus reduces our ability to let our minds wander, which is the basis for observing links which are not immediately obvious. In other words, it's wandering that generates creativity which breaks the functional fixedness. 

You might have observed this in your own life. When you are thinking very hard about an issue, there might be no progress. Nothing new comes to you. Then you do something which you are able to do with very low concentration - for example taking a walk, driving in smooth traffic, going for a swim, taking a shower - and suddenly you get a brainwave. A completely new idea or new perspective comes to you. This is known as transient-hypofrontality you can read more about this at our page on Creativity.

 

(There is a more complicated biological backstory to this, but to appreciate the biology, it would be useful to find understand some basic knowledge on neurons, neural networks, and hormones - in this case, glucocorticoids produced from the stress of trying to finish something quickly).

The Candle Problem is only 1 observation. Do other experiments show the same?

The finding that money as a motivational tool does not work well for creative projects has been quite consistent in many other experiments. 

 

The behavioural scientist Dan Ariely ran an experiment with 87 participants. The participants had to solve a variety of different tasks that were cognitively challenging - requiring memory, creativity, and concentration. The participants came from a rural of India, and were split into 3 groups. Each group was offered a reward for doing well in the tasks, but the reward ranged from small to large. 

edqe.png

What would you expect the results to be? By now, of course, we know better. But when this question was posed to college students, they gave the most intuitive answer - they expect performance to go up as the reward went up. 

 

And the results? The people who were offered the large bonus did worse than the other 2 groups in every single task. The results of the small and medium bonus groups were fairly similar. 

Ariely then designed a similar experiment with students from one the best universities in the world, Massachusetts Institute of Technology. The students had to perform 2 tasks - a cognitive task (adding numbers quickly) and a mechanical task (tapping on keys quickly). Again they were divided into 2 groups, 1 group offered a small reward ($60) for finishing the tasks fastest, while the other was offered a large reward ($600).


The results? What was observed with residents in rural India was the same as highly educated students from a top university.  When the task involved only mechanical skill - bonuses worked as expected. The higher the bonus, the better the performance. But went the task required cognitive skill (and it wasn't even that hard - just addition), a higher bonus led to poorer performance. 

The implication for work

What these experiments show is that monetary motivators are useful in some situations - when the task is straightforward and we know what the solution is. But when the task is complex, requiring thought, strategy, and creativity, and when the solution is not obvious, monetary rewards prove to be detrimental. The anxiety and stress from pursuing this monetary reward causes our minds to narrow, and generally lowers our cognitive performance 

 

But the real issue is not just the one-time performance effect. 

Dr Bernd Irlenbush of London School of Economics analysed 51 experimental studies of financial incentives in employment relations. From his collated findings,  Irlenbush raised another concern, "we find that financial incentives may indeed reduce intrinsic motivation and diminish ethical or other reasons for complying with workplace social norms such as fairness. As a consequence, the provision of incentives can result in a negative impact on overall performance."

In other words, money acts as a contingent motivator in the form of if you want more of X you have to do more of Y - if you want more money, you have to produce more work. 

 

Intuitively, we have become so used to these contingent motivators that this sounds right. But as Irlenbush points out, the cost of this is that people do the work for the money. They don't do work because they think that the work is interesting or worth doing. Moreover, they are more likely to violate workplace harmony in pursuit of the monetary reward, which could come in the form of unethical acts on their colleagues. 

Conclusion


These findings give us plenty to think about. While it might be very difficult to attract talented people without proper monetary rewards, it seems that tying these rewards too closely with performance could have a reverse effect. In pursuing money, people might be less creative, their own individual performance might dip especially for more complex challenges. In the longer term, folks might lose their internal motivation to do good work, and might affect their colleagues or work harmony. 

 

bottom of page