DePaul Law Review

Winter 1998

Symposium:

Contingency Fee Financing of Litigation in America

Third Annual Clifford Seminar on Tort Law and Social Policy

*267 THE WAGES OF RISK: THE RETURNS OF CONTINGENCY FEE LEGAL PRACTICE

Herbert M. Kritzer [FN*]

Copyright © 1998 DePaul University; Herbert M. Kritzer

Introduction

The contingency fee is one of the defining characteristics of civil litigation in the United States. [FN1] The folklore regarding contingency fees produces sharply contrasting images. On the one hand, contingency fees represent the average person's "key to the courthouse." [FN2] Individuals who would otherwise be left with no recourse can pursue *268 the compensation to which they are entitled. [FN3] At the other extreme, contingency fees are the source of all evil in the civil justice system: chasing ambulances for clients, creating expectations that every little injury deserves large payments in compensation, and filing inflated and fraudulent claims. [FN4] As is true of most popular images, there is at least some truth in both extremes. There is no doubt that the contingency fee is extremely important for insuring access to the civil justice system. [FN5] The recent adoption of a kind of contingency fee system in England, long pointed to as a system without such an arrangement, clearly reflects the importance of a workable funding system. [FN6] However, it is also true that the system creates incentives for lawyers to seek out potential clients aggressively, and to find ways of pursuing what sometimes appear to be very dubious claims. [FN7] The fact that the lawyers who stand at the top of the financial success ladder are mostly ones who have made large profits from contingency fee work [FN8] (and that some of these lawyers make no secret of their financial success) inevitably raises questions about profiting from others' misfortunes.

One of the problems for understanding the realities of contingency fees is that most available sources of information provide distorted images. Cases reported in the popular press represent events that are news, not ordinary day- to-day events. [FN9] Jury verdicts reported in the *269 popular press tend to represent extreme outliers, with the average award of the reported cases being ten times or more the average award of all cases. [FN10] Jonathan Harr's recent book, A Civil Action, describes how a big case can create tremendous incentives but also lead to disastrous outcomes for the lawyer who gambles and loses. [FN11] To understand the realities of contingency fee legal practice, one cannot rely upon anecdotal reports of individual cases or lawyers, or on the patterns reflected in news reporting. Only carefully designed, systematic research can provide an accurate picture of typical contingency fee practice patterns and the range and variability of those practices. Drawing upon both extant and new data sources, the analysis that follows describes the economic returns of contingency fee practice. [FN12]

Part I of the paper deals with some preliminaries and then introduces the notion of "effective hourly rate." In Part II, I briefly review two extant studies of effective hourly rates before turning to a variety of sources of information that might provide information on effective hourly rates:

• Unpublished annual income and effort data; and
• The Wisconsin Contingency Fee Study.

• The types of contingency fee arrangements used by Wisconsin lawyers;
• Effective hourly rates on a case-by-case basis (and how those rates vary depending on case, lawyer and practice characteristics); and
• Effective hourly rates based on annual income data.

In Part III, I conclude with a discussion of the implications of my findings for debates over the contingency fee.

*270 I. Understanding Contingency Fees

A. Some Preliminaries

While the most salient difference between contingency fees and arrangements such as an hourly fee or a flat fee is in how the fee is calculated, there are other important differences as well. The normal hourly fee or flat fee simply purchases the services of a lawyer. Under a contingency fee arrangement, the client also purchases additional services. The first is financing. Often clients (particularly individuals) are required to pay for these services up front, before the lawyer performs the services. [FN13] Even if that is not the case, lawyers typically bill their hourly and flat fees, and their expenses, as cases progress, with the expectation, not always correct, that the clients will promptly pay the bill. By their nature, contingency fees are not normally collected until the matter is closed. Very often, lawyers also defer the collection of expenses until the close of a case. Thus, the contingency fee lawyer finances the litigation for the client while a case is pending.

The second additional service that the client purchases is a form of insurance. While in many states clients are liable for expenses regardless of the outcome of a case, the reality is that lawyers who pursue a case unsuccessfully on a contingency basis seldom collect those expenses (or even seek to collect them). Thus, the lawyer effectively insures the client for the expenses associated with pursuing a claim. In addition to insuring for the out-of-pocket expenses, the lawyer also insures the value of his or her time. If the lawyer obtains no recovery for the client, the lawyer absorbs the entire opportunity cost of the time expended on the case. [FN14] In a significant fraction of cases, the lawyer may obtain some fee, but that fee still does not cover the opportunity cost of the lawyer's time. In either of these situations, the lawyer effectively insures the client for the value of the lawyer's time.

The popular image of the contingencies involved in the contingency fee does not fully represent the way the fee works. The obvious contingency is that the lawyer risks receiving nothing if he or she fails to recover for the client. For most contingency fee lawyers this is probably *271 the least important contingency. Rational lawyers want to screen out cases with a low probability of obtaining a recovery. The more important contingencies facing lawyers are the uncertainties over the amount of the recovery and amount of investment by the lawyer that will be necessary to obtain the recovery. At the time a lawyer signs up a new contingency fee client, the lawyer typically has some idea of the general size of the case. However, the reality is that the lawyer only knows that a quadriplegia case is probably worth a lot more than a routine soft tissue case. Cases can change, either increasing in value if the lawyer can locate additional insurance coverage or something goes wrong during the client's treatment, or decreasing in value if the client enjoys an extremely good recovery. In addition, the uncertainties of comparative negligence need to be factored in. How such uncertainties might affect the case depends on the stage at which the case is resolved. [FN15] Even more problematic for the lawyer is that the opposing party's decisions regarding how to handle a case largely determine how much time and effort the case will take. The opposing side might immediately offer to tender its insurance policy limits or, alternatively, fight the case up and down the court system.

B. Understanding Outcomes from the Lawyer's Perspective

The focus of the analysis below is on "effective hourly rates" ("EHR"): the fee received by the lawyer divided by the amount of time the lawyer had to expend to obtain that fee. This measure captures the various elements of the contingencies facing the lawyer. The numerator, the fee received, is a function of both the amount of damages and whether the lawyer obtains any recovery for the client. The denominator is the amount of time the case actually took.

              recovery x damages
 EHR =  -------------------------------
                      hours

Across a number of similar cases, the variation in the numerators and denominators reflect the risks and uncertainties of those cases.

It is useful to focus on this measure because it is precisely what some critics of contingency fees have attacked, suggesting that lawyers are frequently able to obtain "effective hourly rates of thousands and *272 even tens of thousands of dollars." [FN16] While there are some cases that do earn lawyers fees that translate into rates of $1,000 or more per hour, we know little about the frequency of such cases, or, more importantly, what the typical effective hourly rates look like. Economists would argue that the economically rational lawyer would demand to do better, on average, from contingency fees than from hourly (or flat) fees because the contingency fee lawyer is providing additional services to the client which merit compensation. [FN17] However, this type of economic rationality presumes an opportunity cost analysis in which the contingency fee lawyer has alternative uses of his or her time that will provide a known level of compensation. [FN18] In situations where a lawyer has uncommitted time, the lawyer may be willing to accept cases where the lawyer expects the compensation to be less than what the lawyer would like to believe is the value of the time involved. [FN19]

C. Extant Effective Hourly Rate Studies

There is surprisingly little extant research on the kinds of fees and incomes lawyers earn from contingency fee work. Some of the most successful practitioners make no effort to hide their financial success, but these lawyers, persons like Joseph Jamail or John O'Quinn, are not typical contingency fee practitioners. [FN20] Publications targeted at the profession often trumpet the names of the members of the profession with the highest incomes, who are frequently contingency fee practitioners. [FN21] I could find only two published studies that provide systematic information on effective hourly rates earned from contingency fee cases. [FN22]

The first study is from the early 1970s and it considered only medical malpractice cases. [FN23] This study, based on a survey of 671 lawyers, [FN24] *273 reported mean effective hourly rates for plaintiffs' lawyers somewhere in the range of $61 to $84. [FN25] This range primarily reflects uncertainty about and the treatment of co-counsel or referral fees. [FN26] The authors of the study compared these figures to the mean hourly rate charged by medical malpractice defense lawyers, $47, and concluded that the "'effective hourly fee' is not excessively large . . . [and that] plaintiff and defense fees are in the same general ballpark." [FN27] It is worth noting that the use of mean hourly rates here would have tended to skew the figures upward. [FN28] More typically, for data such as these, one would expect median rates to be reported. [FN29]

The second published study on effective hourly rates is from the Civil Litigation Research Project ("CLRP"), a large federally financed study conducted in the early 1980s. [FN30] This study focused on a sample of federal and state cases selected from courts in five federal judicial districts around the country: Eastern Pennsylvania (Philadelphia), Central California (Los Angeles), South Carolina, New Mexico, and Eastern Wisconsin (Milwaukee). [FN31] The analysis of effective hourly rates from this study is reported in the CLRP's Final Report and in The Justice Broker. [FN32] The overall median effective hourly rate for contingency *274 fee lawyers was $42 (the mean was $89). [FN33] In the lowest twenty-five percent of the cases, contingency fee lawyers earned $7 per hour or less. [FN34] In the top twenty-five percent of the cases, contingency fee lawyers earned $94 or more. [FN35] The top ten percent of lawyers in the study earned fees of $198 per hour or more. [FN36] These figures show that very large effective hourly rates are possible, but typical effective hourly rates are not extreme. [FN37] For comparison, the median hourly rate reported by hourly fee lawyers in the same study was $50. [FN38] The information on effective hourly rates from the CLRP study is quite rich. Table 1 shows the information broken down in a variety of ways. [FN39] Two items are particularly interesting. Effective hourly rates tend to go down for small cases (those with awards under $10,000) with a median of $36, and only reach a median range of $57-$58 for the common range of cases yielding awards above $10,000. Also, it appears that contingency fees in cases other than contracts and torts are very risky. For example, in non-contract, non- tort cases, the median is only $6. In fact, thirty-eight percent of these cases yield no fee at all.

There is another way to look at the data in Table 1. This is important because one might be tempted to look at the mean effective hourly rates and conclude that they suggest a problem because of the big gaps between the mean and the median figures. However, this comparison does not give an accurate picture of reality because the variations in the individual case figures depend on both the numerator and denominator of the effective hourly rate equation, and the observed skew might result from one or both of these elements. Because of this, if the interest is in the average fee earned per hour by contingency fee lawyers, we cannot simply average the effective hourly rates. A better procedure is to sum all of the fees earned by contingency fee lawyers and to sum all of the hours worked on those cases by them, *275 and then divide the two. This produces the mean fee earned per hour worked, which is a different figure than the mean effective hourly rate. Doing this yields a mean hourly return of $47 for all contingency fee cases in the sample versus an EHR of $89. Taking only cases under $10,000, which constitute almost half of the sample, [FN40] the mean return per hour is only $25. This is well below the comparable mean of $62.

These two existing studies do not show a large proportion of contingency fee cases yielding extremely high hourly returns for lawyers. The CLRP data show that occasional high hourly returns do occur. [FN41] The situation, however, may have changed in the fifteen to twenty-five years since these studies were completed. It may also be that the frequent *276 high hourly rates come not in cases that get filed in court (which is true of all the cases in the CLRP study, and probably most or all of the cases in the medical malpractice study), but those that are settled quickly before formal litigation begins. In the balance of this Article, I turn to more recent data to see what if anything we can say about typical returns being earned by contingency fee lawyers.

II. Estimating Effective Hourly Rates

Getting reasonable data on the rates charged by lawyers working on an hourly fee basis is reasonably straightforward, and there are many published reports that contain information on these rates. [FN42] However, there is no central registry of cases, settlements, or lawyer effort for cases handled on a contingency basis. Nor are there any regular surveys of the returns that lawyers obtain from contingency fee work. In this Part, I draw upon a variety of sources to get estimates of the kinds of effective hourly rates that lawyers regularly obtain from contingency fee work. None of the sources alone is ideal, but by looking at a variety of sources, it is possible to "triangulate" the information to arrive at reasonable estimates of the rates that lawyers typically obtain. The best information comes from a recent survey that I conducted of contingency fee practitioners in the state of Wisconsin. [FN43] The remainder of this Article devotes substantial attention to reporting analyses of those data. The data have the obvious limitation of reflecting the situation in only a single state, but the patterns are sufficiently similar to what I have obtained from other sources. Therefore, there is a high probability that the general picture the Wisconsin data provides is applicable to many, if not most, areas of the country. As I discuss below, this is borne out by some sketchy results from a 1991 national survey of members at the Association of Trial Lawyers of America ("ATLA"). [FN44]

^aSOURCE: Estimated from various state bar economic surveys (see Appendix 1).

A. Estimating Effective Hourly Rates from Published Data

A number of state bar associations conduct periodic economic surveys of their memberships (see Appendix 1). Some of the reports *277 are sufficiently detailed to permit making estimates of the likely effective hourly rates that typical personal injury contingency fee lawyers are able to achieve. This requires three pieces of information: annual income by specialization after expenses but before taxes, percentage of gross receipts consumed by overhead, and number of hours devoted to remunerative work. Using these data one can estimate a range in which the typical effective hourly rate is likely to fall. I have done this by identifying high and low figures for each component (income, overhead, and hours) from the various state reports, and then using those figures to estimate a likely range. The reports I examined were from Colorado (1993), Florida (1984), Iowa (1990), Michigan (1991, 1994), Nebraska (1994), New Hampshire (1990), Ohio (1990, 1994), Texas (1992), and Wisconsin (1986, 1992). [FN45] One problem with these reports is that there is no common definition of who is a personal injury specialist. In some states, as low as twenty-five percent of the practice is devoted to personal injury work, for others it is as high as eighty percent, and for still others it is simply the area with the largest proportion of the respondent's practice regardless of percentage. [FN46]

These reports show median incomes for personal injury plaintiffs' lawyers ranging from a low of $63,000 in New Hampshire (1990) to a high of $108,000 in Texas (1992). [FN47] Based on these figures, plus the others available, my best estimate of the median income of personal injury attorneys is $70,000 to $80,000. Median figures for overhead as a percentage of gross receipts ranged from thirty-eight percent in Nebraska (1994) to a high of fifty percent in Wisconsin (1992). [FN48] I estimate that the national median is somewhere in the range of forty to forty-five percent. Most of the surveys reported the number of billable hours per week, with figures ranging from a low of thirty- two hours per week in Nebraska (1994) [FN49] to a high of forty hours per week *278 in Michigan (1990). [FN50] If one assumes that lawyers work forty- eight to fifty weeks a year, this produces a range from 1,536 to 2,000 billable hours. For my purposes, I will assume a range of 1,500 to 2,000, with a best estimate figure of 1,650-1,700. [FN51]

• the 90% interval estimate is $69-$88;
• the 95% interval estimate is $65-$112; and
• the 99% interval estimate is $49-$150.

*279

If I had to choose a single point estimate based on this analysis, it would be an effective hourly rate of $80.

This figure may seem low given the popular image that plaintiffs' lawyers are doing considerably better these days than are lawyers defending personal injury ("PI") cases and the assumption that hourly rates for PI defense lawyers are over $100 per hour. However, the economic surveys suggest a murkier picture. [FN53] Table 3 shows median incomes and hourly rates reported in a variety of recent economic surveys. The table includes figures for PI plaintiff, PI defense, general *280 litigation, [FN54] and all lawyers. The hourly rates shown are those quoted in the Iowa and Wisconsin surveys. [FN55] There is no appreciable difference in other surveys (Colorado, Ohio, and Michigan). [FN56] For those surveys where figures were available for general litigators, their incomes tended to fall between figures reported for PI plaintiff and defense attorneys. Note that my best estimate of the effective hourly rate for plaintiffs' lawyers is in the same ball park as the hourly rates quoted by defense lawyers--$75 to $95, except for Texas where the rate is $110. [FN57]

These data also provide some idea of how the relative incomes of personal injury contingency fee lawyers have changed over the years relative to personal injury defense lawyers. This is relevant given the claim that incomes of plaintiffs' lawyers have been rising disproportionately to those of defense lawyers. [FN58] I obtained reports from a number of older economic surveys, five dating from the 1960s (Oregon, Illinois, Pennsylvania, Michigan, and Texas) and one from the early 1970s (Michigan), that broke median income down by specialty. [FN59] Table 4 summarizes the results from these older surveys. [FN60] Overall, these figures do not provide a consistent picture of a significant shift from a situation of income equality between plaintiffs' and defense lawyers. They do seem to suggest that typical plaintiffs' lawyers in large cities have moved from a highly marginalized practice setting, consistent with Carlin's description of the early 1960s, [FN61] to a situation that does not differ sharply from that of defense lawyers in terms of income.

*281

B. Unpublished Annual Income and Effort Data

As part of my research, I obtained the raw data from the 1992 Wisconsin Bar's economic survey. [FN62] This survey included data on primary billing method, income, [FN63] overhead as a percentage of income, weekly billable hours, and number of hours of fee-producing or billable work *282 during 1992. As with the aggregate figures considered in the previous section, these data can be analyzed to produce estimates of the effective hourly rate. This involves inflating the net income to account for overhead [FN64] and dividing by billable hours. The sample is small here because only forty-three respondents indicated that their primary billing method was contingency fee, and only thirty-nine had all of the information needed to estimate the effective hourly rate.

The median effective hourly rate obtained from my calculation was $113, with a range, excluding one extreme outlier, from $18 to $306. The midspread was $70 to $208. The overall median EHR can be compared to the median set hourly rates that lawyers reported, $105. The range of hourly rates was $30 to $275, with a midspread of $90 to $130. [FN65] Thus, the typical effective hourly rates reported by contingency fee lawyers in Wisconsin do not differ sharply from rates charged by lawyers billing on an hourly basis, although as one might expect, there is more variability for the contingency fee lawyers.

• net pre-tax income from law practice from the last tax year (total, broken down by: demographics, legal specialization, race, gender, experience and other);
• average weekly billable hours on legal work (total, broken down by: hourly rate work, flat fee rate, contingency work, and other);
• hourly billing rate;
• the three fields of law which provided the highest source of income during the last tax year, ranked from one to three;
• an estimate of the total expenses per lawyer from the last fiscal year; and

*283

• an estimate of the gross receipts per lawyer from the last fiscal year. [FN67]

Obtaining estimates based on these data was more complex than analyzing the Wisconsin data because most of the lawyers who broke the information down by fee arrangement and who reported spending time on contingency fee work also spent time on work billed on some other basis. I have focused primarily on key results of the Michigan and Ohio surveys. Details of the analysis of the surveys are found in Appendix 3.

The clearest set of figures is for that group of lawyers whose practices probably consist mostly or solely of contingency fee work. I based this determination either on reported time breakdowns or the substance of their practice. I included lawyers who reported that all of their "billable" time was devoted to contingency fee work (their total number of billable hours was equal to the number of hours on contingency fee work) plus those lawyers who did not break their time down by billing method, but listed only contingency fee type work among the three most important areas of practice. [FN68] The number of respondents was small, forty-one for Michigan and thirty-four for Ohio. The range of estimated effective hourly rates for the combined set, from $16 [FN69] to $1,670, [FN70] once again emphasizes the variability, but the medians for the two groups of lawyers, $81 (Michigan) and $88 (Ohio), fall right in the general range of figures I found working from published data. Furthermore, using the combined set of seventy-five respondents, and estimating the total gross income (i.e., inflating their reported net incomes by the same overhead factor used above) and total billable hours, one gets a total gross income of $14,151,120 and a total of 127,280 billable hours. Dividing these figures yields an average return per hour of $111, which is virtually identical to the median hourly rate reported above for Wisconsin. [FN71]

*284 C. The Wisconsin Contingency Fee Study

• a structured survey of contingency fee practitioners to obtain basic descriptive information about the lawyers' practices and information on a sample of cases handled by the lawyers;
• observation of lawyers at work to obtain an in-depth understanding of key processes such as case screening and negotiation; and
• semi-structured interviewing to ascertain whether the observational findings are sui generis.

The analysis that follows draws primarily from the survey, however, because I rely somewhat upon all of the above sources of data, I describe all of them in this section.

1. Data Collection

The survey of contingency fee practitioners, which was carried out during the fall of 1995, relied upon a sampling frame defined by the Litigation Section of the State Bar of Wisconsin. [FN72] Lawyers provided a total of 511 usable responses representing an estimated response rate of forty-eight percent. [FN73] To obtain information on a sample of actual cases, the survey requested data on up to three cases: the case closed most recently after a trial had begun, the case closed most recently after filing but before the start of trial, and the case closed most recently before filing. The "most recent" strategy provides an approximation to random sampling, and the three different disposition stages provide for stratification along the key dimension of when a case is closed. [FN74] Overall, lawyers provided information on 989 cases (332 unfiled, *285 390 filed but not tried, and 267 that went to trial). These cases provide the basis for most of the analyses reported below.

My observations in law offices during 1996 involved three different practices. [FN75] I was excluded from very little that was relevant to my work. [FN76] The three settings were very different. One was a specialist plaintiffs' firm, one was a contingency fee plaintiffs' specialist in a medium- sized general practice firm, and the other was a "litigation" (broadly defined to include criminal, civil and family) specialist in a small general practice firm.

Finally, I conducted a total of forty-seven semi-structured interviews, twenty-eight with contingency fee practitioners, thirteen with litigation defense lawyers, and six with current or retired insurance claims adjusters. I conducted the interviews between May and October, 1996. The sample of contingency fee practitioners was drawn using a combination of legal directories and yellow page advertisements. These interviews averaged about one hour in length, and all were tape recorded and transcribed. The defense side respondents were identified from directories and in the course of interviews with other respondents. [FN77] These interviews were conducted by telephone, and were also, with one exception, tape recorded and transcribed.

2. Characteristics of Contingency Fees

The fraction "one-third" typically comes to mind when the topic of contingency fees comes up. One-third is the "standard" contingency fee figure. [FN78] One reads of fees exceeding this amount, [FN79] and for some types of cases fees below this are either dictated by law (e.g., social security disability cases, workers' compensation cases, and, in Wisconsin, medical malpractice cases), or appear to reflect market forces. [FN80] *286 One of the questions asked on the survey was the exact nature of the contingency fee specified in the retainer agreement.

Excluding those types of cases for which fees are specifically governed by statutes or regulations, 58% of the cases in my sample involved retainers specifying a fee as a flat percentage of the recovery, 39% employed a variable percentage, and 3% employed some other type of contingency arrangement (see Figure 1). Of the cases with a fixed percentage, a contingency fee of 33% was by far the most common, accounting for 92% of those cases. Five percent of the cases called for fees of 25% or less, 2% specified fees around 30%, and 1% specified fees exceeding 33%. Thus, only a little over half of the cases (53%) employed the standard one-third contingency fee. [FN81] This figure is based only on those cases where the attorney had the leeway to charge a one-third fee. [FN82]

The most common pattern for those cases employing a variable percentage called for a contingency fee of 25% if the case did not involve substantial trial preparation (or, in some cases, did not get to trial), and 33% if the case got beyond this point. The contingency fee rose to 40% or more if the case resulted in an appeal. For cases not involving a lawsuit, the contingency fee percentage could be as low as 15% or as high as 33%. The range for those cases involving a suit but not trial was 20% to 43%. For those going to trial, the range was from 25% to as high as 50%. One of the lawyers told me that he would consider taking certain types of risky cases which he saw as having a high likelihood of going to trial only if the contingency fee percentage was 50% if the case went to trial. Another lawyer explained that he would consider quoting a fee that might involve a percentage as high as 50% in cases where the potential client came in with an offer in hand. In these cases, the fee would be based only on any recovery over and above the offer in hand, with the fee being the lesser of 50% of the additional recovery, or 33% of the total recovery. Interestingly, these types of variable fees can work to both the client's and lawyer's advantage. For the client, it means keeping more of the recovery. For the lawyer, it can provide a powerful tool to convince a client to accept

*287

*288 a settlement because the client can actually net more on a smaller settlement where the fee is only 25% and which avoids incurring costs associated with expert witnesses (e.g., pretrial depositions, trial preparation, and actual appearance at trial). [FN83]

• An hourly fee up until an initial settlement offer is obtained, and then fifty percent of anything over and above that offer.
• Hourly fee capped by thirty-three percent of the recovery.
• A flat retainer plus a percentage.
• An hourly retainer plus a percentage once that time is exhausted.
• Hourly fee up to a set maximum with a percentage if that is exceeded.
• Premium hourly rate with no fee if there is no recovery.
• Reduced hourly rate plus a bonus based on recovery.
• Reduced hourly rate plus a reduced percentage.
• Capped hourly rate plus a percentage.

In my interviews, it was clear that some lawyers were very open to negotiating individualized retainer agreements, while others were very firm in offering only specific types of arrangement. Some lawyers expressed a willingness to negotiate with the client to get a case that they viewed as good; others rejected any idea of such negotiations. Others told me that they specifically laid out the choice of an hourly versus a contingency fee. Another lawyer, whose practice was exclusively contingency fee, told me that in a case of clear liability, severe injury, and a relatively low policy limit, he would charge five percent or less (e.g., $5,000 on a $100,000 recovery) if he was able to get the insurer quickly to tender its policy limits. [FN84]

While the fee is usually described as being based on the gross recovery (i.e., before the lawyer is reimbursed for expenses), some lawyers in Wisconsin treat the gross recovery for fee computation purposes as the recovery less any payments to subrogated interests. Even when they do not do this, lawyers typically seek to get the subrogated parties to take a reduced payment which serves as a way of netting more *289 for the client (or as a way of having the subrogated party pay a share of the attorney's fee).

It is not at all uncommon for lawyers to collect fees that are less than what they are entitled to under the retainer agreement. In the survey, I asked lawyers if the final fee differed from the fee specified in the retainer. In 18% of the cases for which the respondents obtained some recovery for their clients, the final fee was less than what they could have taken under the terms of the contingency fee agreement. The survey did not include questions as to why these reductions occurred. Follow-up interviews suggested that two primary elements drove the decision to take a lower fee. First, there was a perception on the part of the lawyer that taking a smaller fee would facilitate a settlement. For example, a lawyer might feel that the client would be more likely to go along if the legal fee was cut from 33% to 30%, or 25% or even 20%. A large proportion of the reductions were from one "round" figure (e.g., 33% or 25%) to another (e.g., 30% or 25% or 20%). Second, some lawyers expressed the view that the lawyer should not walk away with more than the client. In cases in which substantial payments had to be made to subrogated parties, lawyers often reduced their fee to a level that they split what was left after paying the subrogated claims with the client. Occasionally, when the case yields a minimal payoff, the lawyer will simply waive any fees owed. Sometimes a lawyer will waive a fee on a small case as a means of generating good will, particularly if the client is in a good position to refer future potential clients to the lawyer. [FN85]

What this survey shows, particularly when combined with observations and follow-up interviews, is that there is substantial flexibility in the contingency fee system. Clients potentially have a range of options to choose from. This does not mean that potential clients have any idea that there may be alternatives available, or that they "shop" for the best arrangement. It also does not necessarily mean that the "cheapest" is the best; a client clearly would do better to pay a lawyer who can get a $100,000 settlement a thirty-three percent fee than to pay twenty-five percent to a lawyer who gets only a $50,000 settlement. Interestingly, there was no appreciable evidence that the lower percentages were charged by lawyers who were less successful. [FN86] The basic problem for potential clients is a lack of information. Word of *290 mouth provides some information on who is and is not a "good" lawyer, but there is virtually no information on differences in fees. Advertising would provide little information on the quality of service and/or result, but it could provide some very useful information on fee options.

3. Effective Hourly Rates

How well do contingency fee lawyers do in terms of the effective hourly rates they earn from contingency fee legal practice? Recall that by "effective hourly rate" I mean the per hour return for the time that a lawyer devotes to a case. [FN87] Estimating this presents some problems for the contingency fee lawyer for at least two reasons. First, many contingency fee lawyers do not maintain time records for their work on contingency fee cases. Interestingly, the majority of the lawyers who responded to my survey reported that they did keep time records, [FN88] but only about a quarter of the respondents actually consulted their files. Even if all of the lawyers did keep time records, and did consult those records, my observations of the lawyers at work (two of whom did keep time records) made clear that the nature of contingency fee practice (i.e., constant shifting from one case to another) makes tracking time at best an effort at approximation. (This same problem may apply to many hourly fee lawyers as well). The result is that it is typically necessary to rely upon estimates of effort. This means that a specific figure for an individual case might involve some significant error, but if the errors are essentially random, they will cancel out across a set of cases.

The second problem is that if we are going to consider the effective hourly rates obtained by contingency fee lawyers, we need to compare them to something. The obvious comparison is to the fees charged by hourly fee lawyers. In making such comparisons, however, we have to be careful to exclude from the fees obtained by contingency fee lawyers components that would typically be billed separately by hourly fee lawyers. Under both fee arrangements expenses such as copying, travel, witnesses, and filing fees would be handled as separately billable expenses. In contrast, while most hourly fee lawyers would also bill separately for paralegal time, this is an expense absorbed within the typical contingency fee. Consequently, to estimate the effective hourly rate of contingency fee lawyers, we need to deduct from the *291 gross fee the equivalent of what would be charged for any paralegal time devoted to the case.

As suggested above, describing contingency fees needs to be done in some context. One could imagine many possible comparisons. For example, what types of effective hourly rates do various types of physicians earn? I recently had a minor dermatological procedure carried out. The fees by the physician came to $195 for ten to fifteen minutes of his time (and the clinic billed another $112 for the use of its facilities). The hourly rate then was something between $800 and $1,200. Alternatively, one might compare contingency fees to the effective hourly rate charged by a respectable automotive service operator. The stated hourly rate for the mechanic might be $45. However, a customer is billed for the "book time" and a good mechanic can beat the book time by twenty-five to fifty percent. To this rate, one needs to add the mark-up on the parts that the shop sells to its customers. Altogether, a respectable auto mechanic shop might generate $75-$100 per mechanic hour, excluding the wholesale value of the replacement parts.

Another potential comparison is to the hourly rates charged by lawyers with comparable training and experience. I noted above the hourly rates reported by insurance defense lawyers in the economic surveys of state bars. These tend to be in the $80-$100 per hour range. However, this probably understates the actual rate for two reasons. First, insurance defense lawyers bill for everything, including preparing the bill. Second, insurance defense lawyers increasingly rely upon unit billing, which means that they use something akin to a book rate for certain activities. That is, a lawyer might charge out the cost of dictating a letter as not less than twenty or thirty minutes, even if it takes only two or three. Similarly, a phone call may be billed as a minimum of six minutes (1/10 of an hour) even if it took only thirty seconds. [FN89]

Probably the best comparison would be to the hourly rates actually charged by the lawyers who responded to my survey. As it turns out, most of the lawyers (eighty-five percent) had done at least some work on an hourly basis during the previous year. In my survey, I asked them for the hourly rate quoted for the most recent matter they accepted on an hourly basis. A total of 389 lawyers provided information on that hourly rate. The median hourly rate was $125 per hour *292 and the mean was $124. This then provides a baseline for comparison in the discussion that follows.

a. Typical Effective Hourly Rates on a Case-by-Case Basis

There are many features one might explore regarding effective hourly rates on a case-by-case basis: How high can they go? How much do they vary? What is the "typical" hourly rate that a lawyer earns? How does that vary by type of case? The answers to the first two questions are easy: lawyers can, on occasion, earn very high effective hourly rates from contingency fee work. [FN90] I was able to compute an effective hourly rate for 878 cases. About four percent of these exceeded $1,000, and one percent exceeded $2,000. In three of the cases, the EHR exceeded $3,000 with the highest single rate at $4,473. In contrast, in about eleven percent of the cases the effective hourly rate was negative or zero; one lawyer had an effective hourly rate of -$2,617 and another's rate was -$1,225. Thus, if one uses $1,000 as the "jackpot," lawyers were two and one-half times as likely to be total losers than they were to win the jackpot. [FN91] A final indicator of the variability is that the standard deviation for effective hourly rate is extremely high, 429, reflecting the fact that the distribution in effective hourly rates is highly skewed toward a small number of very large figures.

The variability, and the potential of "jackpots," is not surprising. That is, in one sense, the essence of the contingency fee. However, how well do lawyers do in the typical case? How we define "typical" becomes important. The presence of a small number of very high hourly rates leads to the result that we will see very different things depending on whether we look at the median (the middle case) or the arithmetic mean (the common average). In fact, as I argue below, the gap between the median and the mean tells us important things about the nature of contingency fee practice. If I simply take all of the cases in my sample, without considering the lawyer's caseloads or the way I designed the sample (i.e., oversampling cases that went to trial), I find that the median effective hourly rate is $132. This is almost the same as the mean/median hourly rate that these same lawyers report charging for their hourly fee work. [FN92] Thus, in about half the cases in my *293 sample, lawyers did better than the median hourly rate for hourly fee work and in about half the cases they did worse.

If this were the end of the story, an economist would probably conclude that contingency fee lawyers were not pursuing an economically rational course of action, given that the economist expects the contingency fee lawyer to extract higher fees to reflect the risks the lawyer bears and the financing services the lawyer provides. These higher fees appear in the mean effective hourly rate, which is considerably higher, $242. That is, in the typical case, the contingency fee lawyer does not do better than the median hourly rate, but across a set of cases, the lawyer will do much better. This was best expressed by one lawyer interviewed who had a very high volume practice. He stated that sixty to seventy percent of his gross fees came from perhaps a dozen of the cases he closed each year. In most of his cases, he was lucky if he met the costs of running his practice. Eliminating the top ten percent of the cases from the sample leaves the mean effective hourly rate for the remaining ninety percent at $136, which is virtually the same as the overall median. [FN93]

There is yet one other way to look at the information the lawyers provided on individual cases. How much does the average hour devoted to contingency fee work yield? We can estimate this by adding up all of the hours reported on the cases in the sample and all of the fees received (adjusted for the costs of paralegal time), and dividing these two figures. The result, which I will label the "sample-wide mean hourly return," is $169. As with the mean effective hourly rate, this estimate is greatly influenced by relatively small numbers of extremely profitable cases. Dropping the ten percent most profitable cases from the sample leaves a ten percent trimmed sample-wide mean hourly return of $104. Dropping only the top five percent most profitable cases, the mean hourly return is $137, which is virtually identical to the median. This pattern reemphasizes the role of a relatively small portion of cases as generating the "profits" across a portfolio of contingency fee cases.

*294

b. Variations in Effective Hourly Rates

The effective hourly rates might vary systematically based on either case or lawyer factors. Figure 2 is a "box and whisker" plot showing variations in effective hourly rates, both within and across categories of cases. The boxes show the "midspread" of each distribution: the middle half of the cases. The line inside each box denotes the median cases. The lines outside each box extend up to one and a half box length beyond the first and third quartiles. The circles show outliers. Effective hourly rates tend to be higher for cases resolved without filing and for cases involving higher stakes. Effective hourly rates are typically lower for cases going to trial and for cases involving lesser stakes. Interestingly, across types of cases defined by area of law, the median is highest for auto accident cases and lowest for medical malpractice. This actually is more indicative of variations related to risk: risk is lowest in auto accident cases and highest in medical malpractice. The pattern of returns lawyers obtained in medical malpractice cases highlights the relationship between risk and effective hourly rates. One of the lawyers who participated in the observational part of the study was working on a large medical malpractice case. At one point he and I worked through the likely outcomes of the case and their probabilities, which ranged from a fifty percent chance of getting nothing, to a ten percent chance of getting $8 million. We estimated that his "expected" fee was $500,000 (although his actual fee could range as high as $1.7 million under the rules governing legal fees in medical malpractice cases in Wisconsin [FN94]). Given the amount of time *295 the lawyer had devoted to the case, and what was yet to come, I estimated that while he might end up making as much as $1,100 per hour, his expected effective hourly rate was $330. When I later examined the medical malpractice cases in the sample from my survey, I had information on forty cases. The median effective hourly rate was only $38, which is what is shown in Figure 2. However, this reflected, in part, that forty-five percent resulted in no payment at all. The maximum effective hourly rate reported was $2,900, and ten percent of the cases had effective hourly rates of $1,000 or more. The mean effective hourly rate was $334, and the aggregate effective hourly rate across all forty cases (obtained by adding all of the fees and hours across the cases and dividing by forty) was $338 per hour. Clearly, medical malpractice work produces a very good return on the investment of the lawyer's time, but a lawyer faces some very substantial risks in undertaking this work.

*296

Figure 3 shows the relationship between effective hourly rates and several different lawyer-related factors. Generally, the differences on the lawyer dimensions are less than on the case dimensions, but some are still worth noting. First, the women lawyers in the sample appeared to generate effective hourly rates substantially below those generated by men; the median for cases reported by men is $135 compared to $100 for cases reported by women. I have no explanation for this pattern, although it may reflect experience or position in a firm (partners generate a slightly higher median effective hourly rate than do solo practitioners or non-partners). There is little difference in the median depending on the location of the practice. The impact of experience is mildly positive with the medians moving consistently (although only slightly) higher as number of years in practice increases. Not surprisingly, higher overall income is associated with higher median effective hourly rates.

*297

*295 c. Adjusting for Disposition and Case Volume

In some ways the overall figures presented in the preceding section are misleading. Recall that each lawyer provided information about three cases: one unfiled, one filed but not tried, and one tried. The result is that the sample greatly over-represents cases that went to trial, and somewhat over- represents cases that were filed and not tried. Of the 882 cases for which I can compute effective hourly rates, 27% went to "trial" (defined to include administrative hearings in social security cases and workers compensation cases), 39% were filed but did not go to trial, and 34% were not filed. We typically talk about 10% or fewer filed cases being resolved by trial, although cases *296 can go to trial and be settled and cases can be resolved authoritatively by an adjudicator short of a full trial. [FN95] In my sample, because of the way it was designed, 41% of the filed cases "went to trial." This distinction between where in the process a case is closed is important in estimating typical effective hourly rates. That rate is lowest in tried cases (median $96, mean $197). It rises to $144 (median) and $251 (mean) in filed but untried cases, and to $166 (median) and $286 (mean) in unfiled cases. Furthermore, 17% of the unfiled cases yield effective hourly rates of $500 or more, [FN96] compared to 15% for filed but not tried cases, and only 9% of tried cases. Clearly, some adjustments are needed to accurately reflect differences arising from how cases are disposed. To deal with this problem, I developed and applied a set of case weights to adjust the sample to more accurately reflect the distribution of dispositions (see Appendix 4).

A second adjustment is needed to take into account the fact that each lawyer provided information on up to three cases regardless of *297 the number of cases that lawyer disposed of during the time period of interest. Figure 4 shows the effective hourly rate controlling for the lawyer's case volume. In terms of medians, there is a break between those handling ten or fewer cases (median a little over $110) and those handling more than ten (median in the $140-$155 range). What is perhaps more important than the medians is the way that the distributions stretch to the right as the case volume goes up: the higher volume lawyers tend to get a larger proportion of more lucrative cases. For all groups of lawyers, the 25th percentile is around $60. In contrast, the 75th percentile is around $200 for those handling 10 or fewer cases, $250 for those handling 10-25 cases, $330 for those handling 26-100 cases, and $450 for those handling more than 100 cases. The means for the five groups of lawyers are $186, $206, $198, $287, and $392. [FN97]

*299

• the median goes from $132 to $154;
• the mean goes from $242 to $267 (the trimmed mean goes from $136 to $162); and
• the sample-wide mean hourly return goes from $169 to $198 (10% trimmed from $104 to $124, 5% trimmed from $137 to $161).

• the median increases to $171;
• the mean to $317 (trimmed mean to $171); and
• the sample-wide mean hourly return to $221 (10% trimmed to $161, 5% trimmed to $190).

Some skeptics might wonder whether the figures for Wisconsin lawyers are lower than what would be true nationally. A 1991 national survey of members of the ATLA produces figures consistent with what I found in Wisconsin. [FN99] From the available reports from the ATLA study, it is only possible to compute a sample-wide mean hourly return, which is $247. [FN100] This is about ten percent higher than the Wisconsin figure, but it does not involve any adjustment for the cost of paralegal time. In the Wisconsin survey, the adjustment for paralegal costs lowered the fee an average of eight percent. If the same figure applied to ATLA members, the adjusted sample-wide mean hourly return would be $228 (compared to $221 for the Wisconsin respondents).

The adjustments discussed above make it clear that there are obviously profits to be made from contingency fee work. While it is the top ten percent of cases that tend to produce the significant profits, *299 the typical contingency fee practitioner can expect even the remaining ninety percent of cases to produce fee premiums over a portfolio of cases amounting to twenty- five to thirty percent over what hourly fee work generates. Contingency fee work can be very lucrative, particularly for those lawyers who develop expertise and processes for handling large numbers of cases. The high profitability comes from locating a small segment of the cases that produce extremely good returns on the lawyers' investment of time. Some lawyers are able to "cherry pick" the good cases; others handle large volumes of cases in order to find the occasional very profitable case. Relatively few lawyers ever see "the really big one." One of the lawyers observed as part of the study had been doing plaintiffs' contingent fee work for twenty years, had a very successful practice, and had never collected a fee of over $100,000 on a case.

d. Estimating Effective Hourly Rates from Annual Income Data

Recall that I used a variety of sources of extant annual income data to derive estimates of the hourly rates that lawyers generated from their contingency fee work. The survey of Wisconsin contingency fee practitioners included questions that permitted similar estimates. *300 Lawyers provided information on annual income, [FN101] the percentage of their time devoted to contingency fee work, the percentage of their gross fees derived from contingency fee work, the number of hours during the year devoted to work on behalf of clients (as opposed to office overhead tasks and professional activities), and the percentage of the gross practice revenue going to cover overhead. Using the responses to these questions to arrive at an estimate of typical effective hourly rates requires caution. In particular, some lawyers reported devoting a very low number of hours to work for specific clients, which suggests that these respondents probably misunderstood the question. For some of these respondents, taking the data at face value results in figures that appear to be implausible (e.g., a maximum effective hourly rate of $306,818). There are two alternative approaches to resolving these problems: exclude all respondents reporting fewer than some arbitrary number of billable hours (e.g., 500 or 1,000), or arbitrarily impose some minimum on the number of billable hours. I applied both approaches. I should note, however, that these adjustments primarily influenced the estimates of the mean effective hourly rate; the median changed relatively little with the adjustments. I derived two sets of estimates based on annual income data, one including time devoted to screening prospective contingency fee clients and one not including that aspect of the lawyers' work.

Table 5 shows the results of this analysis. The medians from the time on actual contingency fee cases are very similar to what I obtained working from the case level data before adjusting for case volume. Given that the pre- adjustment case-level figures give some measure of effective hourly rate by lawyer (rather than by case), one should expect that these figures are roughly consistent. Interestingly, the figures in Table 5 indicate that the medians fall by about ten to twenty percent when the time devoted to case screening is taken into account. The means are much more sensitive to the adjustments for the problems in the data. The highest estimate, obtained by using a minimum of 500 hours of "billable" time, is roughly consistent with the mean obtained from the case-level data ($346 here compared to $317 previously).

*301

*300 e. The Record Keeping Issue

One of the possible problems with the estimates above is that, even though many of the attorneys in the sample did keep time records, *301 only a small fraction of those who had such records referred to their records in responding to the survey. One might expect that attorneys overestimate their time, either remembering it incorrectly or responding strategically in order to make their per hour return look more acceptable.

When I first thought about conducting the Wisconsin Contingency Fee Study, I had the impression that virtually no lawyers working on a contingency fee basis maintained time records. In conversations with several local attorneys, I became aware that there were at least some lawyers who did keep track of their time while doing work on a contingency fee basis. Drawing upon a list of attorneys who were likely to be in practices which required them to track their contingency fee time, [FN102] I conducted an unscientific survey. I asked these attorneys to provide me with information on contingency fee cases closed over a recent time period. [FN103] These lawyers provided me with information on a total of ninety-two cases (with gross fees received ranging from $0 to $910,000 and lawyer effort ranging from three hours to 7,000 hours). [FN104] As before, dividing net fee by lawyer hours produced an *302 estimate of the effective hourly rate. The median was $125; [FN105] the mean effective hourly rate was $189. [FN106]

In the sample from the scientific survey, there were 151 cases with information on effective hourly rates for which the lawyers reported having consulted their case files and that those files contained time records. [FN107] This represents only seventeen percent of the entire sample used to conduct the Wisconsin Contingency Fee Study and, consequently, the data needs to be treated with caution. For these 151 cases, the median effective hourly rate was $111 and the mean was $170. Looking separately at the unfiled, filed, and tried cases, the respective medians/means are $146/$224 (n=51), $109/$170 (n=61), and $95/$99 (n=39).

Taken together, both the earlier unscientific sample and the sub- sample from the Wisconsin survey show that, if anything, the absence of time records may have lead to an overestimation of the effective hourly rates that lawyers are earning from contingency fee work.

4. Summary: The Returns of Contingency Practice

Not one of the many sources of information discussed in this section can be described as definitive. Critics can find something to challenge in each of the sources that were drawn upon. However, what is striking is that despite their weaknesses, the various sources and approaches point to the same general conclusions:

• The returns from contingency fee practice are at best "somewhat" better than what lawyers earn from hourly fee practices.
• Some, perhaps much, of the surplus disappears when one takes into account the time and effort contingency fee lawyers devote to screening cases.
• A small segment of cases produces substantial "profits," but few lawyers are able to tap into this segment of cases on a routine basis.

The general consistency in the patterns I have described provides a substantial degree of confidence that my analysis presents an accurate picture of routine contingency fee legal practice.

*303 I would, of course, like to have even better data than what are now available. The approach I have pursued with the Wisconsin contingency fee is the method most likely to provide accurate information on routine contingency fee practice.

III. Conclusion: Praying for Justice or Preying on Justice?

Contingency fee legal practice can produce substantial financial rewards. My analysis shows that lawyers who take a case on a contingency fee basis can expect to do somewhat better than they do on their hourly fee cases. [FN108] The occasional very good case will yield some significant profits over and above the balancing out of the typical better than average cases and the worse than average cases. Under certain circumstances, lawyers can do much better from contingency fee work than most lawyers do from hourly fee work.

One way to understand how this works is to think about private legal practice as managing a portfolio of investments. A lawyer whose practice consists entirely of hourly fee work is like an investor who puts his or her money into a bank certificate of deposit. The principal is safe and the return is known in advance. A lawyer working on a fixed fee basis is akin to playing the bond market; there is more volatility, but the amount at risk is somewhat controlled, and there is some expectation of a payoff. The situation changes when contingency fee cases are added to the portfolio.

Most of the lawyers who do contingency fee work also do substantial amounts of hourly fee work. In my sample, only twenty-five percent of the respondents obtained as much as seventy-five percent of their income from contingency fee work; fifty percent of the sample obtained thirty percent or less of their income from contingency fee work. For the lawyers whose practices are predominantly non-contingency fee, the bulk of their portfolios are in CDs (i.e., the returns are quite predictable and the investment is relatively safe). These lawyers invest a portion of their portfolios into risky contingency fee cases with the hope of getting a higher return. In this portion of their portfolios they get a better return than the whole portfolio, but occasionally they do much worse. For example, the mixed practice lawyer I observed had taken a large contingency fee case to trial and lost in the prior year. The resulting write-off had major consequences for the firm's profitability for that year.

*304 The nature of the portfolio strategies of the lawyers concentrating on contingency fee work are somewhat different and varied. A very small group of firms are able to establish reputations that permit them to be extremely selective in the cases they pursue. They can choose either cases that are virtual sure things or cases that have very high potential payoffs. One contingency fee lawyer from outside Wisconsin described his practice as involving a mix of class action cases that involve high risks but extremely high returns, and more routine cases that keep the lights burning. Having the more routine cases permits the firm to devote some of its resources to the high risk, high return cases. This "mixed" strategy is similar to the diversified investor who focuses on very selective blue chip investing with a few occasional speculations.

A variant of this can be thought of in terms of the prices of stocks. Some stocks are quite expensive on a per share basis, and this serves as a barrier to entry for the smaller investor. A small investor might be willing to play the market in terms of a diversified strategy mixing investments of varying degrees of risk and return, but may not have the price of entry to do this with certain types of investments. This smaller investor can still mix degrees of risk and returns. One lawyer, whose practice is primarily social security disability cases, described precisely this to me. He viewed his social security cases as relatively predictable: you win most, lose some, but you know what you have at stake, both in terms of investment of time and possible returns. These cases provided a secure base for his practice. He also took some personal injury cases which involved greater risk and uncertainty but had the potential of higher returns. He felt able to take these risks because of the stable base obtained from his social security practice.

A different strategy for the contingency fee specialist is to try to produce a substantial flow of cases in order to be able to get an occasional "hit." The idea here is that most cases average out between those producing a return above the market hourly rate and those producing a return below that hourly rate. The lawyer is looking for the small segment of cases that can be extremely profitable, either because that segment produces a high fee or because it produces a quick fee. The problem for the lawyer is that he or she cannot be sure which of the cases walking in the door will be these profitable ones. This type of portfolio might be seen as akin to a broadly spread portfolio with a moderate amount of risk. Most of the returns work out to be somewhat better than the savings account/hourly fee type returns. A few of the investments pay off handsomely (i.e., the occasional company *305 that becomes the target of a hostile takeover and pushes the price above what anyone had previously expected).

Finally, there is the lawyer who is more speculative in his or her work. Many of the cases in the portfolio involve high risk, but when the lawyer succeeds the payoffs are very substantial. One might look at this person as a bit like a venture capitalist who regularly engages in taking big risks and knows that a significant portion of the investments will fall flat. However, the successful investments yield handsome returns.

Clearly, contingency fee legal practice can be very good for those who engage in it, but what about the clients' perspectives? Are the lawyers doing well by doing good or are they preying on the misfortunes of others? As with any type of money making situation, greed can be a dominant factor, at least for some actors. Lawyers (and clients) are as susceptible to temptation as others. The fact that some people make a lot of money does not mean that the system has not benefitted others.

• How will the other side respond to a demand?
• What will be the client's damages?
• What are the ambiguities in the factual situation?

Lawyers who do not rely primarily on contingency fee cases may in fact be inclined to advise potential clients that a case would be better handled on an hourly basis. This suggests that there are cases in which the risks are minimal. However, what would a lawyer do if he or she advised a client to rely upon an hourly fee arrangement only to see the case turn sour? I suspect that this has not happened for the lawyers who have had a client use an hourly arrangement on the lawyer's advice. Interestingly, most of the lawyers who told me that they had recommended to a client that a matter be handled on an hourly basis *306 found that clients almost always opted for the contingency fee when advised of even the slightest possibility of a downside risk. Occasionally, potential clients would themselves raise the possibility of an hourly fee. However, most of them were in fact seeking a contingent hourly fee, not a straight hourly fee.

My conversations with lawyers who depend mostly or exclusively on contingency fee cases indicated that they find the idea of handling cases on a traditional hourly basis much more problematic. First, they need the lucrative cases to balance off the cases that they lose or for which they obtain small returns. Recommending to clients with the best cases that they use an hourly fee changes the overall practice equation in ways that are unattractive. Second, some of them are not set up to bill hourly. Third, they view the big profits yielded by some cases as compensation for their willingness to fund cases and take risks. One lawyer told me about a case which came to him on an hourly fee basis. The client had originally retained a lawyer on a contingency basis, but came to the decision that an hourly fee would be a better deal. The first lawyer the client retained worked only on a contingency fee basis and refused to renegotiate the fee. The lawyer and client agreed to part ways, and the lawyer I was speaking to accepted the case on an hourly basis. As it turns out, the case had a lot of problems, which the contingency fee lawyer was willing to deal with but which the client was reluctant to pay the hourly fee lawyer to handle. In a sense, the client was not necessarily in the best position to assess the risks involved in his case.

One certainly hears about clients who are unhappy about how much they had to pay their lawyers. Unhappiness about the cost of legal services is neither a new issue nor an issue limited to clients who paid lawyers on a contingency fee basis. However, there is no indication that those who pay their lawyers on a contingency basis are more likely to be unhappy than those who pay their lawyer on some other basis (e.g., fixed fee, hourly, etc.). Increasingly, corporations are trying to move toward contingency arrangements with their attorneys. The fact that some of the most successful contingency fee attorneys are those who handle very large cases on behalf of major corporations indicates that the most knowledgeable consumers of legal services are willing to pay purely on a results achieved basis even if that means astronomical hourly rates in the end. [FN109]

*307 In some ways, the bottom-line question, and probably the most difficult question to answer, is whether clients are ultimately better off or worse off given the contingency fee as opposed to alternative mechanisms of paying for legal services? Many contingency fee clients probably do pay more for legal services than they might if they paid by the hour. However, many of those same clients would probably never seek redress if it were not for the insurance function provided by the contingency fee. In a sense, clients pay a premium for eased access to the civil justice system. Furthermore, many, perhaps most, clients are able to have access precisely because of the availability of a system like the contingency fee. In a fundamental sense there is a tradeoff between access and cost, where the access issue is a combination of risk shifting from the client to the attorney and the availability of funds up front to purchase a needed service.

Could one design a system which would permit access but reduce the amount contingency fee clients pay? Recent proposals have suggested that percentages be strictly controlled in cases where the client receives an early offer. As noted above, in Wisconsin, many attorneys charge lower percentages for cases resolved earlier in the process. [FN110] In a sense, the market has, at least in Wisconsin, already made this available. Critics might view even these lower fees as too high. Perhaps they are in some cases, but designing a mechanistic system which would allow attorneys and clients to determine which cases are generating excessive fees and which are not, is very difficult. The current proposals fail to take into account the uncertainties that typically exist regarding damages, liability, and sources of compensation at the time a potential client first contacts an attorney. None of the systems envision any type of post-hoc review of fees (something which technically is already available under inherent judicial authority). Post-hoc review is potentially time consuming and costly, particularly in the absence of well-established standards, but relying upon such a system would probably serve to avoid the problems of the mechanistic systems others have proposed.

One of the ironies of recent proposals is that they have come from quarters one would expect to be most likely to support market solutions to pricing. One could imagine a couple of market-related alternatives that might serve to bring down existing fees. The first would *308 be to find ways to get more information about fees to potential clients so that clients could make more informed choices. The obvious mechanism for this would be what amounts to price advertising. There is, of course, the problem that cheapest is not necessarily best. However, most consumers are fully aware that there is a price/quality trade-off. Yet, simply making consumers aware that there are alternatives in pricing will put some market pressures on the providers of the service.

Competition could also be opened up by permitting non-lawyer specialists to handle cases within some defined specialties. For example, is there a reason why knowledgeable consumers should not have the option of retaining private insurance adjusters to negotiate on their behalf with an insurer? Particularly in those types of cases where the concern about excessive charges by lawyers is greatest (i.e., the clear liability, policy limits case), someone who knows the insurance claims system should be able to represent a claimant's interests effectively. If Kritzer's Claims Service will do the case for fifteen percent, while Kritzer & Kritzer, S.C. will charge thirty percent, then the consumer should have that choice. [FN111]

Finally, one could imagine a system in which claims were actually sold to lawyers in an auction-like setting. [FN112] Lawyers would bid for claims: the better, more certain the claim, the more the lawyer would offer (i.e., the less the lawyer would discount the projected recovery). Obviously, there is the problem that the full value of a claim is not necessarily known at the time a potential client may first want to retain a lawyer. This might be handled by a system of lawyer brokers, where the broker would sign up the client for a small percentage of the ultimate recovery (i.e., five or ten percent), and provide the initial "front-end" legal assistance (i.e., doing the things necessary to preserve the claim, collect initial evidence, etc.). [FN113] Once the claim had matured, the broker would put the claim up for bid. To avoid the problem of the client walking away once the auction was completed, the amount of the payment could be put into escrow, to be collected only when the claim was actually paid or the case was otherwise resolved. If the claim was potentially lucrative but very risky, clients would have the option of selling the claim at a steep discount, or entering *309 into something that looked like a traditional contingency fee arrangement.

The goal should be to find ways of increasing the options in the existing system and making potential clients aware of those options. Legislative limitations on fees are likely to have more negative than positive consequences. In contrast, while deregulation has risks, competition among service providers is more likely to have long run payoffs than is increased regulation. Typically, limitations on legal fees have served to reduce the availability of legal representation. [FN114] In contrast, creating competitive mechanisms has the potential of not only lowering fees but also of making services available in cases where legal representation was previously unaffordable.

*310 Appendix 1

Economic Surveys Conducted by State Bar Associations

CO 1967 [mentioned in Mich. B.J., May 1974, at 9]

CO 1972 Susan Taylor & Julie Honnold, Report on the 1972 Economic Survey of the Colorado Bar (1973).

CO 1977 Colorado Bar Association, Economics of Law Practice Committee, Report on the 1977 Economic Survey of the Colorado Bar (1977).

CO 1982 1982 Economic Survey of Colorado Lawyers (1984).

CO 1989 Colorado Bar Association 1989 Economic Survey (1992).

CO 1993 1993 Economic Survey of Colorado Attorneys (unpublished statistical summary by Colorado and Denver Bar Associations provided by David Brand).

FL 1984 The 1984 Florida Bar Economics Survey (1984).

FL 1986 [mentioned in Results of the 1992 Annual Survey of Florida Bar Members: Economics & Law Office Management 53]

FL 1989 Results of the 1989 Annual Survey of Florida Bar Members: Economics & Law Office Management (1989)

FL 1992 Results of the 1992 Annual Survey of Florida Bar Members: Economics & Law Office Management (1992)

GA 1968 [mentioned in Ga. St. B.J., Aug. 1969, at 112]

IA 1955 Iowa State Bar Association, Incomes of Iowa Lawyers; A Preliminary Report (1956).

IA 1968 [mentioned in Mich. B.J., May 1974, at 9]

IA 1980 Philip A. Houle, The 1980 Economic Survey: Iowa State Bar Association (1982).

IA 1985 Philip A. Houle, The 1985 Economic Survey: Iowa State Bar Association (1985).

IA 1990 Philip A. Houle, The 1990 Economic Survey: Iowa State Bar Association (1990).

IL 1959 Daniel J. Cantor, The Economic Picture Today . . . The Illinois Survey and Spot Check of Representative Law Offices, 49 Ill. B.J. 884 (1960).

*311 IL 1964 Billie R. Bethel, Report of the 1964 Economic Survey of the Legal Profession of Illinois (1964).

IL 1971 George A.M. Heroux, Results of the 1971 Illinois Bar Economics Survey, 60 Ill. B.J. 182 (1971).

MI 1963 The Economics of Law Practice, 43 Mich. B.J. 11 (1964).

MI 1973 The Economics of Law Practice in Michigan, 53 Mich. B.J. 1 (1974).

MI 1981 John M. Wright, 1981 Economics of Law Practice Survey of Michigan Lawyers, 61 Mich. B.J. 20-23, 112-19 (1982).

MI 1984 John M. Wright, 1984 State Bar of Michigan Economics Survey, 64 Mich. B.J. 1186-1190, 1306-1312, 65 Mich. B.J. 22-33 (1986).

MI 1988 State Bar of Michigan, Legal Economics Section, The 1988 Desktop Reference on the Economics of Law Practice in Michigan (1988).

MI 1991 The 1991 Desktop Reference on the Economics of Law Practice in Michigan (1991).

MI 1994 The 1994 Desktop Reference on the Economics of Law Practice in Michigan (1994).

MS 1970 [mentioned in Mich. B.J., May 1974, at 9]

MO 1958 Benchmarks for Your Practice, 15 J. Mo. B. 12 (1960).

MO 1962 Where Are You on the Totem Pole?, 19 J. Mo. B. 17 (1963).

MO 1966 [see brief description of the study in Gerald B. Rowan, Economics of the Bar, 23 J. Mo. B. 576 (1967)]

MO 1970 Wade F. Baker, 1970 Survey: Salaries and Non-Salaried Lawyers Income by County Class, Employer, and Age, 28 J. Mo. B. 535-39 (1972).

MO 1974 Cecil L. Caulkins, 1974 Economic Survey--Income of Salaried and Non-Salaried Lawyers, 33 J. Mo. B. 221-27 (1977).

MO 1982 James C. Wirken & Cecil L. Caulkins, 1982 Economic Survey of Missouri Lawyers--An Overview, 39 J. Mo. B. 541-48 (1983).

NB 1994 1994 Membership Survey Results, NBSA News, June 1994, at 5-11 (Nebraska State Bar Association).

NC 1994 North Carolina Bar Association Economic Survey 1994.

*312 NH 1990 The 1990 Desktop Reference on the Economics of Law Practice in New Hampshire (1990).

NJ 1962 Daniel J. Cantor & Co., The 1965 Economic Survey of Pennsylvania Lawyers: Conducted by the Committee on Economics of Law Practice of the Pennsylvania Bar Association 20 (1965).

NJ 1988 Economic Survey of New Jersey Lawyers (1988).

NM 1982 1982 Economic Survey of New Mexico Lawyers (1982).

NM 1989 1989 Economic Survey of New Mexico Lawyers and Law Firms (1990).

NY 1960 [mentioned in Daniel J. Cantor & Co., The 1965 Economic Survey of Pennsylvania Lawyers: Conducted by the Committee on Economics of Law Practice of the Pennsylvania Bar Association 23]

OH 1952 Milton Kafoglis, Economic Condition of the Legal Profession in Ohio, 28 Ohio B. 866, 907, 935, 978 (1955).

OH 1990 Ohio State Bar Association, The 1990 Desktop Reference on the Economics of Law Practice in Ohio (1990).

OH 1994 The 1994 Desktop Reference on the Economics of Law Practice in Ohio (1994).

OK 1970 [mentioned in Mich. B.J., May 1974, at 9]

OR 1963 David A. Baerncopf & Donald A. Dole, Survey of the Economics of Law Practice (1963).

OR 1971 [mentioned in Mich. B.J., May 1974, at 9]

OR 1981 [mentioned in Economics of Law Practice: 1989 Survey Results (1989)]

OR 1989 Economics of Law Practice: 1989 Survey Results (1989).

PA 1965 Daniel J. Cantor & Co., The 1965 Economic Survey of Pennsylvania Lawyers: Conducted by the Committee on Economics of Law Practice of the Pennsylvania Bar Association (1965).

TN 1971 [mentioned in Mich. B.J., May 1974, at 9]

TN 1994 The Tennessee Bar Association Statewide Survey of Law Firm Management (1994).

TX 1967 Texas Economic Survey, 31 Tex. B.J. 6 (1968).

TX 1987 The State Bar of Texas 1987 Membership Survey (1987).

*313 TX 1989 Hourly Rate Report: The State Bar of Texas 1989 Law Firm Economics and Management Survey (1989).

TX 1992 Attorney Compensation Report: The State Bar of Texas 1992 Attorney Billing and Compensation Survey (1992).

TX 1992 Hourly Rate Report: The State Bar of Texas 1992 Attorney Billing and Compensation Survey (1992).

US 1988 Altman & Weil, Inc., The Small Law Firm Economic Survey (1988).

WA 1974 The Economics of Law Practice in Washington State (1974).

WI 1983 Frank Murphy, State Bar Membership Survey: Where the Big Bucks Aren't--A Look at Lawyer Income, 57 Wis. B. Bull. 28-29 (1984).

WI 1987 Survey Reveals: Wisconsin Attorney Income Up 11 Percent, 61 Wis. B. Bull. 14-16 (1988).

WI 1992 Becky Weiner, 1992 Economics of Practice Survey Report (1993).

*314 Appendix 2

Estimating Effective Hourly Rate From Annual Income Data

In this Appendix, I detail the construction of the results shown in Table 2. Table A shows the same data contained in Table 2, but each row is shown with a row number and each column with a column letter. Each cell in the table is designated by its column/row position.

Rows 1, 2, and 4 were taken from the various economic surveys from 1990 through 1994. No effort was made to adjust Row 1 for inflation, because it is not clear to me that there has been that much change over the period, or if there has been a change, that it is consistent with standard price/income inflators/deflators.

Row 3 was computed by dividing Row 1 by 1 minus Row 2. That is, cell A3=A1/ (1-A2) or $98,000=$63,000/(1-.36).

Row 6 simply divided Row 3 by Row 4. That is, cell A6=A3/A4 or 65=98,000/1,500.

Row 5 is the tricky one. Here I divided the highest income (D3) by the lowest hours (A4), and the lowest income (A3) by the highest income (D4), etc. Specifically:

A5 = A3/D4    or      49 = 98,000/2,000

  B5 = B3/C4    or     69 = 117,000/1,700

  C5 = C3/B4    or     88 = 145,000/1,650

  D5 = D3/A4    or    150 = 225,000/1,500

TABLE A

COMPUTING THE ESTIMATES OF EFFECTIVE HOURLY RATES

FROM ANNUAL INCOME DATA^a

	A low estimate	B best low estimate	C best high estimate	D high estimate
1 Annual net income before taxes	A1	B1	C1	D1
2 Overhead as a percent of gross	A2	B2	C2	D2
3 Estimated gross income	A3	B3	C3	D3
4 Billable hours worked	A4	B4	C4	D4
5 Effective Hourly Rate 1	A5	B5	C5	D5
6 Effective Hourly Rate 2	A6	B6	C6	D6

*315 Appendix 3

Analyzing the Michigan and Ohio Lawyer Economic Surveys

As mentioned in the body of this paper, the data from the Michigan and Ohio economic surveys raised a number of analytical problems. Theoretically, the information the respondents provided should have allowed me to focus in on the contingency fee effort. That is, assuming the simple case of a lawyer who does both hourly and contingency work, I could take the lawyer's estimated gross, multiply time billed on an hourly basis by hourly rate and subtract that out from the gross. This calculation should leave the income generated by contingency fee work which could then be divided by the hours on contingency fee work to provide an estimate of effective hourly rate.