Countries often fail to reach an agreement in negotiations because the negotiators themselves are selfish or uncooperative. A prime example are U.N. summits on climate change or the recent refugee crisis in the E.U. where countries can’t seem to agree to a fair quota to accept the migrants. One reason high stakes negotiations fail time and time again is because negotiators, often elected by us to represent our interests, are extortionists. Researchers from the Max Planck Society and Harvard University who applied an intricate game theory model that simulates social dilemmas faced by negotiators found  extortioners keep their own contribution to a collective target to a minimum. Moreover, 40% of participants resorted to extortion and when they didn’t want to extort, participants elected someone else to do the dirty work.

Machiavellian cooperation

Extortionists can actually achieve a collective target, but not if there’s too many of them in the group. Image: Pixabay

In a setting where cooperation is required, otherwise all players lose, extortionists seem to win the most. An extortionist will cooperate occasionally and in this way entices his co-players to cooperate more often as this is the only way that the co-player can increase his modest gains.

For the first time, researchers tested the outcomes of extortionist behaviour in large groups. Inspired by climate change negotiations, the researchers recruited 630 students from the Universities of Bonn, Hamburg, Göttingen, Kiel and Münster and had them play a climate game three times. Each participant was awarded 40 euros to invest in climate change protection. The participant could choose to invest anything from the whole sum to nothing, and keep any money left over. The catch is that a target had to be reached consisting of each participant donating half the cash on average (20 euros),  otherwise everyone would lose everything. Quite akin to the realities of climate change.

The participants were divided into 15 groups of 18 players each. Groups were then sub-divided into 6 ‘countries’ of 3 players each who elect, re-elect or vote out their representative after each game. The elected representative would then have to prevent climate change (the object of the game) over 10 rounds. Each player was able to campaign for the role of the negotiator in the second and third games. Generally, these campaigns involved either selfish promises (“we’ll contribute less than others, so we’ll have more money leftover”) or cooperative ones (“we’ll make sure we succeed and make our fair contribution”). A representative could be voted out by two compatriots based on the behaviour in the previous game and be replaced by another citizen of the ‘country’ in the following game. As a control, the researchers also assembled group of eighteen and six students each who directly and without elected representatives could make their own endowments.

(a) The 6-players treatment, (b) the 18-players treatment, (c) the 6-representatives treatment. Each game consists of 10 rounds, during which players need to raise sufficient contributions to reach a specified target sum. Games 2 and 3 are replicates of game 1. The players remain the same in the 6-players and the 18-players treatment. In the 6-representatives treatment, representatives are randomly picked in game 1 and re-elected or voted out for games 2 and 3. Re-election of a representative may depend on the representatives’ performance in previous games. In addition, except for the first four groups, after games 1 and 2 all players in the 6-representatives treatment are asked to write non-binding pledges about how they would contribute if elected. Players are only informed about the pledges of members of their own subgroup.

Group in which six players directly contributed to save the climate reached the target more often than those game where six elected representatives paid the contributions on behalf of their electorate. Groups involving 18 independent contributors, however, performed worse than the representatives at negotiations. This was expected since individuals lose sight of a collective purpose as the parent group becomes larger. This is why representatives are better at achieving a collective target, but only if they don’t fall in the same pitfall: becoming integrated in a large negotiating group.  “So G8 is better than G20,” says Manfred Milinski from the Max Planck Institute for Evolutionary Biology.

Many of the representatives actually met the climate target. That’s impressive, but that’s only because fair representatives had to compensate for the deficit caused by extorting behaviour. Forty percent of the players engaged in extortionate behaviour during the climate game. Moreover, those candidates who openly promised to contribute less than the fair share were likelier to get elected. “As a result, there were more unfair players among the representatives than among the six independent decision-makers. This was what the electorate wanted,” explains Milinski.

 “In this way, they were all winners: the fair representatives and their ‘compatriots’ who elected them to a lesser extent, and the unfair representatives and their compatriots to a much greaterextent,” says Jochem Marotzke from the Max Planck Institute for Meteorology in Hamburg.

“We repeatedly observed that players consciously decided not to resort to extortion,” says Milinski. The opposite applies, however, to the representatives: “Psychologists believe representatives behave differently when they have to represent other people. It would appear that they unconsciously avail of this latent potential so that they can fulfil the expectations of their clientele.”

It seems that while most of us do not like to behave like extortionists, we choose to elect those that are extortionists for us. What’s more, this study seems to prove that Machiavellian cooperation works: all parties involved win at the end, it’s just that extortionists benefit more, while fair parties benefit less. Chances are no one will ever win, though, in a room packed with extortionists.

Findings appeared in Nature Communications.