The simple model of litigation spending can be used to analyse the effect of cost-shifting rules if parties go to trial. A cost-shifting or fee- shifting arrangement simply specifies who pays which costs, depending on the result of the trial. There are two main cost-shifting systems that have been analysed in the literature:
- • The American rule - each party pays their own costs, irrespective of the result.
- • The English rule - the loser pays all of the winner's costs.
It is important to remember that even when the English rule applies, it is almost never the case in practice that the loser pays all of the winner's legal costs. Institutions have evolved to limit the kinds of costs that can be claimed. Typically, after a legal trial concludes, the successful party files its bill of costs (which sets out the winner's claim) with the court. An officer of the court then assesses the 'reasonableness' of the costs with reference to legal precedent and a statutory schedule of limits of entitlements of costs. The level of reduction usually means the bill of costs is reduced by between 50 per cent and 70 per cent, depending upon the jurisdiction, and the 'reasonableness' of the bill of costs in the first place. This process is known as the 'taxation' of legal costs. This 'taxation' is not related to the ability of the government to impose commodity or income taxes - the 'taxation' of costs is simply the process of examining - and, where necessary, reducing - the bill of costs of a lawyer. 'Costs' include not only the lawyer's own professional fees, but also the disbursements incurred. Thus, one of the purposes of 'taxation' is to determine the amount of costs a successful party in litigation is entitled to recover from their unsuccessful opponents.
One of the main results in the literature is that aggregate spending is higher under the English rule than it is under the American rule.2 Intuitively, the American rule reduces the payoff from winning, and increases the payoff from losing, relative to the English rule. The American rule therefore reduces the gap between winning and losing, thereby lowering the overall 'stakes' involved in the case. Thus, less effort is devoted towards trying to win the case (and towards avoiding losing the case) than it otherwise would be. This means that the expected surplus 'in the shadow of the court' is lower under the English system and higher under the American system. Thus, although the American rule may involve lower trial expenditure, it may create incentives for vexatious plaintiffs to file frivolous lawsuits, because the expected surplus from filing a lawsuit is higher. Thus, one prediction is that the English system creates higher costs per trial, but the American system encourages more lawsuits.
The effects of cost-shifting rules on trial expenditures can be easily modelled within our framework. Suppose that the loser of a trial is obligated to pay a fraction S < 1 of the winner's costs. Let P be the per unit price of employing a lawyer. Then the expected payoffs are:
for the plaintiff and
for the defendant. The first-order conditions in this game yield: for the plaintiff, and for the defendant.
The left-hand side of each of these expressions is the marginal benefit of litigation effort. The right-hand side is the marginal cost of litigation effort or hiring a lawyer. Note that when S = 1 (which is the pure English rule), marginal benefit always exceeds marginal cost, and so there is no well-defined best response for either player. Thus, when S = 1 - the pure form of the English rule - there is no pure strategy Nash equilibrium.3
However, as soon as S < 1, marginal benefits are diminishing and may reach unity (as long as eP and eD are sufficiently high), so best responses are well defined. These two equations imply that eP = eD. Substituting in to player 1's first-order condition, we get:
The expected payoffs are:
Note that these payoffs are both decreasing in S and can be negative, in which case the parties would both be better off if they could agree to set eP = eD = 0. However, this is not an equilibrium: if the plaintiff agreed to put in no effort, the defendant could choose some small positive effort level of ? and obtain an expected payoff of:
whereas if he put in zero effort his payoff would be:
Since ? can be made arbitrarily small, it is never a best response to put in no effort. This means that there is no level of S > 0 which induces the parties to choose the same level of expenditure as under the American rule.