2007_02_01_archive

Congo meets Game Theory

I know as about as much about game theory as I do about auto

mechanics, but I read an article about the prisoner's dilemma and it

stuck with me. By contrast, I am of the opinion that if the oil really

needed changing, the car would simply stop.

In the prisoner's dilemma, two accomplices are arrested for a crime.

The police officers interview them individually. The men have no

opportunity to communicate with one another. The police offer each

suspect the same choice: he can receive immunity and testify against

his accomplice, who will be given a ten-year sentence. Alternatively,

he can stay silent. If the men stay silent, both receive a six-month

sentence on a minor charge. Both men betraying one another would

result in a two year sentence for each man. The choices boil down to

cooperating with the partner or defecting. The dilemma is that neither

man knows for certain what choice the other will make.

The solution that maximizes their total welfare is if they both

cooperate and do not betray one another. They serve their six months

and then are free to pursue their nefarious activities or write a

tell-all book about prison life. However, each man knows that the

other has a personal incentive to defect and rat him out. No matter

what the other person does, a suspect will gain by defecting. If his

(obviously now former) friend did not do the same, the suspect will

get no jail time. If his partner also defected, then at least he only

serves two years as opposed to ten. The ground-breaking, Nobel-prize

winning conclusion is that self-interest will motivate people to

defect (this of course was later contradicted by ground-breaking,

Nobel-prize winning rebuttal demonstrating motivation to cooperate).

Even with no walls and cells, people in Congo assume that others will

cheat the system and defect. They therefore defect as well, even at

the cost of collective welfare. In this context it makes sense to

maintain your army even if you are running for office. Your opponent

is doing the same. Disbanding them now might leave you holding the bag

with the proverbial ten-year sentence.

The cops do not trust the government to send paychecks. Neither do the

soldiers. In this dilemma, they actually are damn near certain the

government will defect rather than cooperate, so they had best do the

same to minimize their loss by preying on civilians. The people

through whom government money passes each take a cut because everyone

else does. The only actions people can depend on in Congo are ones

driven by self-interest. Cops harassing civilians on the street,

governors commandeering NGO vehicles, people stealing bridge

materials, generals giving false information, staff pilfering from

employers...these are not the problems. They are certainly problems,

but they are not THE problem. They are symptoms of the overall

problem: an absence of trust causing mass defection rather than

cooperation, at the expense of collective welfare.

The elections in Congo are no magic bullet precisely because the

prisoner's dilemma will still exist. It will continue until people

have faith that others will not try to cheat the system. A democratic

government will only be effective to the extent that defectors are

punished and cooperators rewarded. People must trust one another. I

should really look up that theory on coorperation, since apparently,

it would have the answer.