In 1949 Harry Harlow, a professor of psychology, carried out a series of experiments observing rhesus monkeys solving a puzzle which looked similar to the latch on a door with a pin securing it. At the start of the period of experiments he placed the puzzle in the cage with the monkey and waited for them to get used to their new set of surroundings. What happened next caught him completely by surprise.
Without any inducement at all the monkeys set to work trying to solve the puzzles. They worked with what seemed like diligence and looked to be enjoying the task. As the days went by the monkeys became more adept and quick at solving the puzzles until they were opening the latch-like mechanism in no time at all.
This went against the grain of what was known about what powered behaviour. Harlow knew that there were two main motivators in life – the biological and the external. The first is present in all animals, the need for survival – to eat, drink and copulate to produce offspring. The second drive, he knew was an external one, that of reward or punishment from the environment based on our actions. In the case of the puzzle solving monkeys, neither of these could have been motivators for them as ‘the solution did not lead to food, water or sex gratification’, nor was there a reward for solving or punishment for not solving the puzzle.
Harlow realised that there must be a third drive taking hold of these monkeys, compelling them to solve the puzzles provided them. ‘The performance of the task provided intrinsic reward’, he wrote. The monkeys seemed to be solving the puzzles because they found it gratifying, they enjoyed the challenge of doing so. He called this Intrinsic Motivation.
In one of his final experiments in this series, he decided to test the strength of this new third drive as surely it must have been subordinate to the biological and extrinsic motivators. So, he experimented with rewarding the monkeys with raisins whenever they solved the puzzles, what could be more gratifying for a monkey than to receive food for doing good work. However, what occurred shocked Harlow even more. The monkeys actually got WORSE at solving the puzzles, making more errors and completing them less frequently. It seemed that providing an external inducement decreased performance. (1)
Who decided that paying a performance bonus was a good idea anyway?
In 1900, Henry Winslow Taylor demonstrated his ‘Scientific Management’ method of producing more, faster with less material at the Exposition Universelle in Paris. Having turned down a place at Harvard to go to work in a steel factory he had become fascinated by how such a revolutionary technology such as the day’s finest industrial machinery could be utilised in such a haphazard, unscientific way by the people working there. Methods of carrying out processes in the factory were a hangover from the days of artisans, knowledge was passed down through the team and instructions were given using approximations at best. He felt that there must be something to be done to improve the way work was performed.
By applying reductionist thinking and using a stop watch to measure his results, he broke down the actions of the factory workers into their constituent parts and set about devising methods of reducing the amount of time it took to complete them. The small reductions in the time it took to perform each task in his new method of working, meant that the finished product was formed in a vastly reduced time overall.
The usual rate of manufacture employed in factories like Taylor’s cut nine feet of steel per minute. His new method cut fifty feet, it was nearly 6 times faster. The Marginal Gains much trumpeted by Team Sky and David Brailsford are not a new concept.
At the exposition in Paris, his demonstration caused a storm. People queued for hours to catch a glimpse of his operation. Once news spread, they travelled across Europe to see for themselves this miraculous way of churning out product. Prominent Industrialists of the time wrote ‘Nobody quite believed at first in the prodigious result..but we had to accept the evidence of our eyes’, the process was ‘a landmark in the history of mankind’ and likened the breakthrough to the invention of the electric lightbulb.
Taylor’s Scientific Management was so popular that people devoted their lives to his vision. It moved from factory to factory, industry to industry, always improving what had gone before. It meant that workers became cogs in a machine built for efficiency. It also meant that what was once skilled work became simple steps that could be executed by anyone once trained. What once were seen as complex tasks only to be performed by an educated worker could now be carried out at a more efficient rate by someone unskilled, uneducated and cheaper. To ensure these unskilled and uneducated workers worked in the right way, at the right time, producing as much as the system could achieve; methods of reward and punishment were brought in to have more of the behaviour that was required and less of the behaviour that was not. Eventually, this settled on a monetary incentive to work. If a worked produced what the system said they could they were paid accordingly, if they over achieved this target they would be paid more. Thus was born the Bonus Scheme.
This premise of industry, to reward the good work and punish the bad, has remained in our everyday lives since that demonstration in the 1900 Parisian display. It has worked well. Very well. It has helped us transform our work from blue collar to white collar – there has been a best way to insert paper into a typewriter, the quickest way of clipping paper together and scripts for handling sales calls. Scientific Management and the financial incentive has built our economy to where we are today. However, what has worked before has now to all intents and purposes become a hindrance.
As economic progress has shifted from blue collar to white, Taylor’s reductionist methods could be applied to reduce tasks to scripts, formulas and step-by-step processes. Scientific Management is in effect today with, what was once high skilled, highly educated white collar work carried out in the Developed World of North America, Western Europe, Japan, South Korea and Australia moving off-shore to where it can be completed cheaper.
Whilst outsourcing the routine work has become more and more prevalent and efficient to do, the type of work the Western World produces is becoming more complex. The work that North America, Europe, Japan etc have begun to produce is that of problem solving, learning and discovery, what Daniel Kahneman, the author of Thinking Fast and Slow, has termed Heuristic Work. McKinsey & Co have estimated that 70% of job growth in the United States is that of heuristic work (2).
Kahneman says that work can now be divided into two categories – algorithmic and heuristic – the former meaning process oriented, the latter artistic and non-routine.
Algorithmic work can be outsourced and automated – technology can now perform tasks that only a few years ago relied on a human to carry out, for example tax preparation software and automatic retail checkouts. In fact a recent study by PwC warned that 10 million jobs in the UK are set to be replaced by technology in the next 15 years (3). Heuristic work cannot be outsourced or automated.
Whereas the old methods of reward and punishment work to a certain extent in the world of algorithmic work, findings say that they hinder the production of heuristic work upon which Western economies now rely.
Studies show that rewards hinder performance
Harvard Business School’s Teresa Amabile has found that creative heuristic work relies heavily on Harlow’s third drive, intrinsic motivation (4). Subsequent experiments into rewarding heuristic tasks conducted by Edward Deci and colleagues found that there was a ‘hidden cost of rewards’. They observed a group of pre-school children who enjoyed spending their free play time drawing and devised and experiment to test what effect providing a reward had on this activity, which the children clearly enjoyed. The children were divided into three groups – those given no reward for drawing, those given an unexpected reward for drawing and those told that if they spent time drawing they would receive a reward. A few weeks later, when the researchers visited the pre-school again they could clearly see that children who were in the no rewards and un-expected reward group drew with the same enjoyment and relish as before. However, the children that had expected a reward for drawing showed much less interest and drew less than previously. They were experiencing a fact which we all know, that once enjoyable work can become a drudge. However, only those that were expecting a reward for drawing were impacted by this. Those that did not expect a reward yet received one for drawing still had the same verve for the task as before.
Their conclusion was that ‘if-then’ rewards, if a task was performed then a reward would be give, affected the performance of heuristic tasks. The reason they gave was that these ‘if-then’ rewards asked people to forfeit some their autonomy. They varied this experiment over and over with different groups of people and still found the same. Extrinsic rewards, especially ‘if-then’ rewards in over 100 experiments were shown to have a negative impact on intrinsic motivation of heuristic tasks (5).
‘I don’t need my team to be intrinsically motivated’, I hear you say. ‘I pay people for performance of their tasks at work. When they perform well, they get paid more’. But this has been proven to have the reverse effect.
In a 2005 study performed by economists from MIT, Carnegie Mellon and the University of Chicago for the Federal Reserve Bank of Boston, high pay was shown to be detrimental to high performance. In their study of 87 participants in India, they set a series of tasks from unscrambling anagrams to recalling a string of numbers, all requiring creativity, concentration and some requiring motor skill. Again, divided into three groups the participants were to earn a reward. Group 1 would earn 4 rupees (at the time worth around a day’s pay) for hitting their performance target, Group 2 would earn 40 rupees (about two week’s pay) and group 3 would earn 400 rupees (nearly 5 months pay).
It turned out that the medium level of reward had no affect over the level of performance produced by participants than the low level of reward. However, the high level of reward made participants perform worse than the other two groups on nearly every task. ‘In eight of the nine tasks we examined across the three experiments, higher incentives led to worse performance’, they concluded (6).
Another study carried out by the London School of Economics in 2009 analysed 51 studies of corporate pay for performance plans. The conclusion was that financial incentives have ‘a negative impact on overall performance’ (7).
As economists in the previous study in India wrote in their report ‘one cannot assume that introducing or raising incentives always improve performance’. They also stated that using bonus structures in the corporate world may be ‘a losing proposition’.
As these experiments clearly show, extrinsic rewards given in the usual if-then way, such as a pay-for-performance bonus, can have a negative effect on your team. These findings were so controversial that the researchers were forced to re-analyse almost 3 decades of data from their experiments. Still they came to the same conclusion ‘when institutions – families, schools, businesses and athletic programmes, for example – focus on the short-term and opt for controlling people’s behaviour’ they do so to the detriment of long-term performance (8).
- Harlow et al., “Learning Motivated by Manipulation Drive” , Journal of Experimental Psychology 40, 1950
- Bradford et al., “The Next Revolution in Interaction”, McKinsey Quarterly 4, 2005
- Hawksworth et al., “UK Economic Outlook July 2017”, PwC
- Teresa Amabile, “Creativity in Context”, 1996
- Mark Lepper, David Greene and Robert Nisbett, “Undermining Children’s Intrinsic Interest with Extrinsic Rewards”, Journal of Personality and Social Psychology 28, 1973
- Ariely et al., “Large Stakes and Big Mistakes”, Federal Reserve Bank of Boston Working Paper No 05-11, July 2005
- “LSE: When Performance-Related Pay Backfires”, Financial June, 2009
- Deci at al., “A Meta-Analytic Review of Experiements Examining the Effects of Extrinsic Rewards on Intrinsic Motivatioin”, Psychological Bulletin 125, 1999