October 31, 2012
As part of the reorganisation of a library, the participants in the experiment were asked to spend one day searching the shelves for specific books. With regard to the configuration of the field experiment, Goerg explained that, "We wanted the conditions to be as natural as possible, which means the participants were unaware that they were being observed." Goerg and Kube divided their test subjects into various groups whose payment was performance related.
In the reference group, the participants received a basic wage plus a further 10 euro cents per find. Members of the second group received the same remuneration, but as a kind of self-chosen goal they were asked before they started to forecast how many of the books they would find. "If they found fewer books, though, there was no monetary consequence whatsoever," explain the researchers in their recently published working paper. Monetary consequences did apply to the third group, however. They were also asked to forecast their performance in advance, and were rewarded with a bonus if they achieved this self-chosen goal. There were two further groups in the experiment which had to find a specified number of books. The first group had to find 100 books, the second only half as many.
The study showed that the different work goals and types of remuneration had a significant effect on performance and motivation. With an average of 116 books found per person, the performance of the group who chose their own goals was measurably higher than that of the reference group paid a piece rate for each item only. No less interesting to Goerg was the observation that the participants tended to overestimate their productivity if there were no negative consequences for not achieving the self-chosen goals. Even when they realised that they would never achieve their targets, they still worked significantly harder than those with no target. "After all, they knew they would still receive money for each book they found," said the researcher.
As soon as a bonus is involved, however, people are evidently more realistic about estimating their work performance: in this part of the field experiment, 64 percent of the test subjects achieved their goals, compared with only half of those not offered a monetary incentive. When the targets were not self chosen but prescribed by others, this had a negative effect on performance, although the participants did still find higher numbers of books than the reference group without any goal at all.
Therefore, prescribed goals also boost performance. However, it is better to allow each employee to formulate their own goal. "Targets for entire groups of people are problematic because of the differences in individual productivity, and can be counter-productive," warns the economist. "Certain goals could motivate some, yet discourage others." For, as the experiment showed, some people achieve with ease targets that others find extremely hard work.
The study findings are thus in line with previous psychological studies related to the effectiveness of targets. As well as corroborating the importance of targets for success, however, the study also demonstrates that it can be advantageous for employees to choose their personal goals themselves and thus determine the potential financial bonus and the degree of difficulty. "The combination of self-chosen performance targets and financial incentives has a particularly motivating effect on work performance," conclude Goerg and Kube.