Why go to court?
Whilst most legal disputes do settle out of court, we know that there are obviously exceptions to this: in the real world, legal trials do in fact occur. So why might parties go to court? Various explanations have been advanced in the literature.
One obvious answer is asymmetric information: if one party has very reliable information regarding the evidence that will be presented at trial that will help his case and the other does not, then they may both be willing to go to court. To see how asymmetric information can affect the likelihood of trial, suppose that there are two possible types of plaintiffs: those with a high probability of victory at trial (p = ph) and those with a low probability of victory at trial (p = pl), with ph > pl. A plaintiff knows his own probability of victory, so each type of plaintiff can compute his expected gain from going to trial. For convenience, we will assume that there are no costs involved in settling.
For each type of plaintiff, we can compute the expected gain from going to trial. For the high types, the expected value of going to trial is:
For the low types, the expected value of going to trial is:
Assume that defendants cannot observe the plaintiff's true type. The defendant only knows that in general there is a fraction в of high types in the population, and a fraction (1 - в) of low types in the population. If the defendant has incomplete information, he perceives the average plaintiff probability of victory to be p = 0ph + (1 -0)p. Suppose the defendant has incomplete information, and makes a settlement offer of S > 0 to the plaintiff, where this amount of money is comprised of a monetary transfer and the value of the efficient level of product to the plaintiff.
The maximum offer that the defendant would be willing to make to ensure that all types of plaintiffs accept the offer must satisfy:
So setting s = Sh = phU(Q) - CR is the lowest offer the defendant could make to ensure settlement occurs. The defendant's expected net benefits under such a pooling strategy are и (Q) - [p,u(Q) - CR] = (1 - fp) и (Q) + CR .
On the other hand, suppose the defendant follows a separating strategy and makes a settlement offer so that only low types would be willing to accept. Low types would be willing to accept as long as:
is the lowest offer the defendant could make to ensure settlement occurs only with the low-type plaintiffs. The defendant's expected costs of making such an offer are:
Finally, suppose that the defendant makes a settlement offer so that no types of plaintiffs would be willing to accept. This is the always go to trial strategy. Setting S = 0 ensures that the case always goes to trial, and the defendant's expected net benefits of making such an offer are:
Which strategy should the defendant follow? The defendant should follow the strategy which involves the highest expected net benefits. Let us rank these strategies in order of their costs.
First, note that the always go to trial strategy is never the best option, since the expression in (11.16) always exceeds that in (11.17). The separating strategy will therefore have the highest net benefit of all three strategies if:
If в is sufficiently small, then there is a relatively low probability that the defendant faces a highquality plaintiff. This makes it more likely that the lefthand side of (11.18) is less than the righthand side, and makes the separating strategy more attractive. This means that when в is sufficiently small, there will be a chance that cases will go to trial, since the offer in (11.15) will only be accepted by low types, and high types will want to go to trial.
On the other hand, if в is sufficiently large, then there is a relatively high probability that the defendant faces a high-quality plaintiff. This makes it less likely that the left-hand side of (11.18) is less than the right-hand side, and makes the pooling strategy more attractive. In this case, the defendant will not allow any cases to go to trial.