Law & order

Does power lead to corruption?

Scientific research shows that, whatever an individual’s personality type, power leads to antisocial decisions – and testosterone plays an important part too

 Silvio Berlusconi: could the proven ability of power to corrupt the leaders who wield it explain his exploits? Photograph: MARI/SINTESI/SIPA/REX

John Antonakis

Wednesday 17 December 2014 07.00 GMT

A few years ago, I published a paper in the journal Science showing that even small children can predict the results of parliamentary elections with two candidates merely by looking at their photographs. As you might imagine, the result caused quite a stir, and it led to an interesting discussion with a journalist.

After a debate over the presidential elections in the US, a journalist and I began to discuss the case of former Italian prime minister Silvio Berlusconi. The journalist wanted me to explain why Berlusconi was so power-hungry; what factors might explain his character and his sexual escapades?

I did not want to answer. I don’t like to respond on particular cases because it’s hard to know if a single case is indicative of a causal relationship. However, after persistent questioning, I blurted out a response: “I don’t know. Maybe it is because of his testosterone?”

But an answer given out of frustration and in the heat of the moment became prophetic. We know that the hormone testosterone indicates reduced empathy and increased antisocial behavior, as well as controlling sexual behaviour. So when planning a study on leader corruption with my colleagues, I thought it might be interesting to measure testosterone too.

What we wanted to study was the question of whether power really corrupts. We recently published our findings in the Leadership Quarterly.

Does power corrupt?

Answering this question using observational data is not easy. It is possible that power corrupts, but it is also possible that some individuals have a natural inclination to seek power because they are corrupt at heart. We were therefore interested to see if stable traits or dispositions that we could measure, such as personality (for example, honesty) or physiological factors (such as testosterone levels), mattered for corruption.

To know whether power does corrupt we had to exogenously manipulate power; we gave power to a random group of participants and observed how they behaved. By random assignment, we ensured we had roughly equal numbers of similar individuals (honest, smart, corrupt, men, women, and so on) in our experimental and control groups. If we found differences in levels of corruption, the explanation could not therefore be that the groups comprised different types of people at the outset.

We set up two lab experiments. Participants played what is called a dictator game. The dictator, referred to as the “leader” in the experiment, could decide how to apportion a sum of money between himself or herself, and his or her team. They had to make choices between serving the greater good – doing what’s right for public welfare by increasing the team’s payout – or serving oneself, thereby increasing the leader’s payout but destroying public welfare.

We also manipulated the number of followers for each leader: the leaders either had one follower (low-power leaders), or three (high-power leaders). Low-power leaders had few choices with respect to abusing their power and high-power leaders were given more options. We then allowed the leaders to take decisions about payouts.

What did we find?

In the first experiment, results showed that high-power leaders took antisocial decisions at a significantly higher rate than low-power leaders.

The second experiment was more complex: we added a time component and monitored participants’ individual differences as predictors of their behaviour. Prior to becoming leaders or followers, we asked participants to vote on what a responsible leader should do with respect to payouts. Most endorsed the pro-social option; just 3.33% said that leaders should take antisocial decisions.

Yet, when they became leaders, participants succumbed to the corruptive effects of power. Interestingly, honest individuals were initially shielded from taking antisocial decisions – but, with time, even they slid down the slippery, corrupting slope of power. Even more interesting was our observation that those who had high levels of testosterone were most corrupt when they had high power.

Our findings have important implications for the design of institutions. Leaders prefer to have decision-making autonomy and lots of discretion. Yet we know from our study, and others too, that power can go to leaders’ heads.

Perhaps our findings do not explain everything about dominant high-profile individuals such as Robert Mugabe, Silvio Berlusconi, or Dominique Strauss-Kahn, or their contemporaries in the corporate world. But they do suggest that institutions should limit how much their leaders are allowed to sip from the seductive chalice of power.

Source: http://www.theguardian.com/sustainable-business/2014/dec/17/does-power-lead-to-corruption-research-testosterone