هزینه های مشروط در مقابل هزینه های احتمالی: مشوق های هزینه های شکایت های قانونی

Many jurisdictions in the UK and Australia now allow conditional fees to be struck between a lawyer and their client. While both conditional and contingent fees are outcome-contingent, only under the former is the lawyer's remuneration a function of inputs. The public policy arguments put forward for each fee structure hinge importantly on litigation expenditure incentives. We show that conditional (contingent) fees are not better suited to the English (American) rule of fee-shifting. We also show that whether conditional fees result in higher or lower expenditure than contingent fees depends crucially on whether the inputs to litigation are determined by the lawyer or the client. If the latter, then conditional fees result in (weakly) lower expenditure than contingent fees. If the lawyer determines inputs then expenditure is higher under conditional fees.

The passing of the Courts and Legal Services Act, Section 58, in England and Wales in 1990 has resulted in litigants being able to enter ‘conditional fee’ arrangements with their lawyer (Bowles, 1994).1 The Conditional Fee Agreements Order 1995 brought this new system into effect and was substantially revised in 2000. Conditional fee agreements are now generally permissible in all but family and criminal matters (Jeffries, 2002). Similarly, conditional fees are now allowable in the states of New South Wales and Victoria in Australia. Conditional fees are similar to the contingent fees widely used in the US in that both ensure that the lawyer is remunerated only if their client wins—they are outcome-contingent. They differ in that the lawyer obtains a fraction of the plaintiff's award under a contingent fee (often one third), while obtaining a percentage above the market price for their services under a conditional fee (to a maximum of 100% in the UK and 25% in Australia). Our objective is to compare the expenditure incentives associated with these two types of fee structures, with a view to understanding the environments to which each is best suited. The pros and cons of contingent fees have been analyzed extensively (Danzon, 1983; Halpern & Turnbull, 1983; Hay, 1997, Miceli, 1994 and Rickman, 1999; Rubinfeld & Scotchmer, 1993; Schwartz & Mitchell, 1970; Swanson, 1991). In contrast, conditional fees have received relatively little attention to date, a notable exception being the analysis by Gravelle and Waterson (1993). This lack of attention is presumably due to conditional fees being a recent innovation and not having been adopted in the US. Our aim here is to compare the litigation expenditure incentives under the two types of fee. Whether conditional fees, ceteris paribus, result in higher or lower expenditure than contingent fees is important as it directly impacts on whether the underlying public policy goals for introducing conditional fees are likely to be realized. Two important public policy motivations for introducing conditional and contingent fees, rather than fixed fees, are to increase access to justice and to reduce the misalignment of incentives between the lawyer and client in relation to the quantum of legal inputs used. If, for example, conditional fees can be shown to result in higher expenditure than contingent fees, then it can be argued that contingent fees are more effective than conditional fees in addressing the misalignment of lawyer–client incentives. The issue of access to justice has at least two dimensions—the affordability of litigation and the ability to access representation and the court system even when it is affordable. However, the argument that outcome-contingent fees enhance access to justice for litigants with meritorious suits but few resources focus only on the first of these two dimensions (Fenn & McGuire, 1994; Rubinfeld & Scotchmer, 1993).2 We likewise focus only on this dimension here.3 The rationale is that outcome-contingent fees promote affordability by ensuring that litigants with meritorious suits incur no costs in the (unlikely) event they lose at trial, while if they win, their costs are funded from their award for damages. While the above argument holds straightforwardly for contingent fees, it is not so obvious that conditional fees enhance access to justice since the lawyer's fee is not tied directly to the plaintiff's award but rather is a mark-up on the value of services provided. Thus, the effectiveness of conditional fees in enhancing access to justice depends on how they affect litigation expenditure incentives. If litigation expenditure increases significantly under conditional fees, then a winning plaintiff's access to justice may in fact be reduced below that under both fixed and contingent fees. Turning to the second public policy argument, it is widely recognized that contingent fees provide lawyers with stronger incentives for exerting effort than fixed fee agreements. First, ignoring reputational issues, since lawyers have no stake in the trial outcome when contracting on a fixed fee, they have weak incentives to engage in productive effort (i.e. effort that influences the trial outcome). Second, lawyers have an incentive to over-provide unproductive inputs (i.e. inputs that benefit only the lawyer, not the client, such as the hiring of unnecessary staff). This is the problem of ‘over-servicing’ or ‘cost-padding’. 4 Thus, fixed fee agreements provide the wrong incentives with respect to both the amount and the mix of legal inputs. 5 Both fee structures address the first of the above incentive problems, as lawyers have a clear incentive to employ sufficient productive inputs if their remuneration is conditional upon winning. Also, contingent fees clearly address the second issue since if the lawyer's fee is proportional to the damages awarded, then the lawyer bears the full cost of employing unproductive inputs. This eliminates the incentive to over-service. The failure of conditional fees to break the linkage between input use and lawyer fees means that the problem of over-servicing is not directly addressed. We show that conditional fees do indeed result in higher expenditure than contingent fees if the lawyer controls litigation expenditure. However, the opposite is true if the client controls litigation expenditure.6 Thus, the relative welfare properties of the two fee structures depend importantly on who is in control of litigation expenditure. The two different views of who controls litigation expenditure have been widely studied. The client-controlled litigation perspective implicitly assumes that the client is able to perfectly control the lawyer's actions and has sufficient expertise to determine the optimal expenditure level ( Braeutigam, Owen, & Panzar, 1984; Gong & McAfee, 2000; Gould, 1973, Katz, 1988, Landes, 1971, Posner, 1972 and Shavell, 1982). In contrast, the lawyer-controlled litigation perspective stresses the lawyer–client agency problem ( Bebchuk, 1988; Dana & Spier, 1993; Hay, 1997, Png, 1983, Rickman, 1999 and Watts, 1994). While most cases lie somewhere between these two extremes, the precise balance of control between client and lawyer is difficult to measure. In order to understand the range of possible outcomes, both ends of the spectrum are considered. Since contingent fees are outlawed in many jurisdictions in Britain and Australia while conditional fees are absent from the US, it might be conjectured that conditional (contingent) fees are more suited to the British (American) legal environment. Since one of the most striking differences between the two environments is in regard to fee-shifting (from winner to loser), it might be conjectured that conditional (contingent) fees are better suited to the English (American) rule of cost allocation. In order to allow a systematic treatment of this conjecture, we explicitly model fee-shifting. We establish that any comparison of the two fee structures depends not only on whether litigation expenditure is client- or lawyer-controlled, but also on whether the fee and quantum of legal inputs are determined simultaneously or sequentially. If expenditure is client-controlled and simultaneously determined with the fee, then contingent and conditional fees have identical behavioural implications. The client is forced to fully internalize the effect of their expenditure on the lawyer's fee when they are simultaneously determined. The fact that both types of fees are subject to the same competitive forces in equilibrium means that the precise structure of the fee has no impact on the client's optimal expenditure. If the fee is determined first followed by litigation expenditure, then we show that contingent fees result in (weakly) higher expenditure than conditional fees under client-controlled litigation. Moreover, this is independent of the extent of fee-shifting. Thus, the rankings are identical under the American and English rules. It follows that the decision to ban contingent fees but allow conditional fees in Britain cannot be rationalized on the basis that the English fee-shifting provisions apply. Rather, other factors such as divergent views about the strength of various incentive effects must underlie the different choices.7 Fee-shifting also has no impact on the ranking of expenditures under the two fee structures when litigation is lawyer-controlled. The lawyer's remuneration is independent of fee-shifting since this simply determines who pays the lawyer – the plaintiff or the defendant – not whether the lawyer gets paid. Thus, the lawyer's incentive to employ additional legal inputs is independent of the extent to which costs can be shifted, regardless of whether contingent or conditional fees apply. The intuition in relation to client-controlled litigation is more complicated since expenditure is affected by fee-shifting in this case—the explanation is deferred. Gravelle and Waterson (1993) also examine how contingent, conditional, and fixed fee agreements impact upon litigant incentives and dispute outcomes. However, their analysis differs from ours in a number of respects. First, they model the defendant's effort to avoid an accident from occurring but do not model litigation expenditures at trial. Second, they analyze a larger span of the time-line of a dispute, modeling both pre-trial settlement and the filing of suits.8 They also model the degree of altruism of lawyers, and their finding that this is an important factor is consistent with our approach of distinguishing between lawyer- and client-controlled litigation. While focussing on a narrower slice of the dispute resolution process, we derive more explicit comparisons of contingent and conditional fees, thus allowing for a clearer understanding of their relative merits. The model is described in Section 2 followed by a discussion in Section 3 of how contingent and conditional fees are determined in equilibrium. Then the cases of client- and lawyer-controlled litigation are analyzed in Sections 4 and 5, respectively. Conclusions are drawn in Section 6.

We have compared litigation expenditures under contingent fees to that under conditional fees. We find that while expenditure is always (weakly) higher under a contingent fee if the client controls litigation expenditure, the opposite is true under lawyer-controlled litigation. Also, we show that the ranking of expenditures is always independent of the extent of fee-shifting, regardless of who controls expenditure. A natural direction in which to extend this line of inquiry is to formally model the effect of taxation of costs under the English rule since this is an important mechanism for controlling moral hazard. Indeed, this mechanism may adequately compensate for the shortcomings of conditional fees in correcting for the incentive of lawyers to over-service, thus providing insights into why conditional fees may be an appropriate choice under the English rule but not the American rule of fee-shifting. A more detailed model of how fees are determined in equilibrium, particularly in the lawyer-controlled litigation model, could also increase our understanding of the relative merits of contingent and conditional fees. In particular, it could be useful to formally derive optimal contracts from the client's perspective as in Santore and Viard (1999). Another obvious extension is to incorporate an analysis of how the two fee structures affect the incentives for litigants to reach a pre-trial settlement. Not only are most disputes settled out of court, but also by applying the logic of backwards induction, our analysis of the expenditure incentives at trial (and their effect on each party's perception of winning) provides the necessary foundation for such an analysis. However, in order to proceed, it is necessary to impose additional structure on the legal technology so that the precise magnitude of the expenditures, and hence beliefs, of both litigants can be deduced under the two fee types. In the absence of such structure, our ability to draw conclusions about the effect of each fee on pre-trial settlement is very limited. In fact, it is possible only to conclude that if the expenditure of both parties is higher under fee A than under fee B to the same extent, so that their beliefs are the same under fee A and fee B, then pre-trial settlement is more likely under fee A. This follows since they have the same expected benefit from going to trial under each fee (since their beliefs are unchanged), yet the costs of going to trial are higher under fee A, implying that it is less attractive. Since this is clearly a very restrictive set of conditions, determining the relative effects of the two fee types on pre-trial settlement will necessitate specifying the properties of the legal technology in more detail.