In every country, centralized government is funded by mandatory public taxation and that means a layer of bureaucracy to go after tax evaders.

It can obviously be good for the community - people get along better when they aren't micromanaging the finances of police and fire stations on a monthly basis. And in small communities, self-policing is easy. Everyone knows who the cheats are.

In relatively large communities, it is more difficult - individuals do not always know who they are impacting and so some don't obey the rules and often exploit the willingness of others to cooperate.  But what is the crossover point for where a community decides to take its money from direct benefits, police and fire stations, etc., to indirect, like hiring more government to pursue tax fraud?

An interdisciplinary team of researchers used an experimental game to investigate the conditions under which institutions of this kind can arise and found that a group of 'players' does particularly well if it has first used its own tax money to set up a central institution which punishes both free riders and tax evaders - but they only set up institutions to penalize tax evasion if they have decided to do so by a democratic majority decision. 

Who in a democracy is going to solve a problem that doesn't exist? San Francisco is an example of a city always proactively banning things without evidence but rational players do not. The irony is that democracy enables the creation of rules and institutions which, while penalizing individual sacrifice, are best for the group.

What does that mean for a larger problem, like climate change? It means that social authoritarians who want to ban, mitigate and ration are fighting a losing battle. The only way to 
create a path to common climate protection measures is on a foundation of democratic conditions.

The experiment involved allowing participants in a modified public goods game to decide whether to pay taxes towards a policing institution with their starting capital. They were additionally able to pay money into a common pot. The total paid in was then tripled and paid out to all participants. If taxes had been paid beforehand, free riders who did not contribute to the group pot were punished by the police. In the absence of taxation, however, there would be no police and the group would run the risk that no-one would pay into the common pot.

Police punishment of both free riders and tax evaders quickly established cooperative behavior in the experiment. If, however, tax evaders were not punished, the opposite happened and the participants avoided paying taxes. Without policing, there was no longer any incentive to pay into the group pot, so reducing the profits for the group members. Ultimately, each individual thus benefits if tax evaders are punished.

But can participants foresee this development? To find out, the scientists gave the participants a choice: they were now able to choose individually whether they joined a group in which the police also punish tax evaders. Alternatively, they could choose a group in which only those participants who did not pay into the common pot were penalized. Faced with this choice, the majority preferred a community without punishment for tax evaders – with the result that virtually no taxes were paid and, subsequently, that contributions to the group pot also fell.

In a second experimental scenario, the players were instead able to decide by democratic vote whether, for all subsequent rounds, the police should be authorized to punish tax evaders as well as free riders or only free riders. In this case, the players clearly voted for institutions in which tax evaders were also punished. "People are often prepared to impose rules on themselves, but only if they know that these rules apply to everyone," summarizes Christian Hilbe, the lead author of the study. A majority decision ensures that all participants are equally affected by the outcome of the vote. This makes it easier to introduce rules and institutions which, while demanding individual sacrifice, are best for the group.

The participants' profits also demonstrate that majority decisions are better: those groups which were able to choose democratically were more cooperative and so also made greater profits. "Democracy pays – in the truest sense of the word," says senior author Manfred Milinski from the Max Planck Institute for Evolutionary Biology. "More democracy would certainly not go amiss when it comes to the problem of global warming."

Citation: Christian Hilbe, Arne Traulsen, Torsten Röhl, and Manfred Milinski, 'Democratic decisions establish stable authorities that overcome the paradox of second-order punishment', PNAS, 23 December 2013