Lawyers routinely settle lawsuits or advise their clients about settlement. One might expect, therefore, that clients routinely complain about some aspect of their lawyers’ settlement services. Ten years of data from eleven jurisdictions paint a vivid, different picture: although the vast majority of civil lawsuits are resolved through negotiated settlements and although complaints against lawyers are common, fewer than 1 percent of reported legal malpractice cases and only about 1.5 percent of bar complaints relate in any way to lawyers’ settlement-related conduct or advice. Even in those instances when clients do raise such complaints, clients rarely prevail. In short, even though lawyers play a prominent role in settlement, lawyers currently operate with no meaningful exposure to complaints about this important aspect of modern practice.

Why do lawyers enjoy this level of de facto immunity? The current legal malpractice system operates on three basic assumptions about lawyering—each of which contemplates the lawyer’s role in litigation, rather than the lawyer’s role in settlement. First, the law assumes that a lawyer’s strategic judgments should enjoy the highest level of deference. After all, we would not want to second-guess decisions about which witnesses to call or which legal theory to advance at trial. Second, the law assumes that any mistake a lawyer makes will be reflected in dampened prospects in litigation. This permits us to employ the case-within-a-case method to determine whether a lawyer’s misconduct caused the client to lose a case they otherwise would have won. And third, the law assumes that clients’ compensable interests are bounded by the remedial powers of the court. Thus, the appropriate measure of damages for any instance of litigation malpractice is the difference between what the court actually did and what a court would have done if the lawyer had been minimally competent. Each of these assumptions is animated by the limited image of lawyer-as-litigator, and each might be defensible if the malpractice system dealt only with alleged litigation errors.

The challenge is that these same three familiar principles shield lawyers from virtually all accountability in the context of legal negotiations, even though we live in the age of settlement. The current system need not—and should not—persist. Negotiated settlements are here to stay. Lawyers will continue to play important roles in those settlements. Clients should be justified in believing not only that their lawyers are improving at this aspect of their practice, but also that their lawyers are accountable when they fall short.



A lawyer who has been working with her client on a high-stakes piece of litigation for more than a year comes to him1 and says:

Based on all that I now know about your case, I think you have a very slim chance of winning outright, if this goes to trial. Opposing counsel just called, and they have offered to settle at $1 million, along with a confidentiality provision roughly consistent with your goals. Their offer expires today, though, and I see no reason to think we’ll do any better, so I advise you to accept it.

Perhaps the lawyer’s advice is excellent, and she has done a great job in all aspects of this representation. She made good decisions throughout the litigation, in ways that maximized the client’s prospects of winning a difficult case. She negotiated wisely and effectively on behalf of her client, maximizing the value the settlement option will produce for her client. And she has done her best to communicate clearly with her client about the choice he now faces.

Or perhaps the lawyer has been substandard in her representation. Her assessment of the client’s litigation prospects is inaccurate. Or her poor litigation decisions up to this point are the reason the case is now doomed. Or she has so little experience trying cases that she now has stage fright. Or the proposed settlement terms will not, in fact, provide the confidentiality the client seeks. Or the proposed settlement is structured in a way that fails to take advantage of tax-minimizing opportunities that would be obvious to most lawyers. Or the attorney actually received the offer last week but only just now communicated it to the client. Or the “going rate” for cases like this is much more attractive than this offer. Or the attorney already told the other side that this deal would be acceptable to her client.2

Most clients likely believe that their lawyers have provided outstanding service in such contexts—and most clients are likely correct in that assessment. Given that most lawsuits settle3 and that lawyers play a central role in settlement negotiations and decision-making, however, we must assume that at least some lawyers sometimes fall short of the profession’s standards of care in the settlement context. The lawyer-client relationship is marked with information asymmetries and potential incentive misalignments.4 Clients often have a limited ability to assess the wisdom of their lawyers’ advice because lawyers have specialized comparative expertise. Clients have few inexpensive ways to monitor their lawyers’ actions.5 And lawyers are human beings, prone to “screw up”6 once in a while. In short, it must be the case that attorneys sometimes fall down in the context of settlement lawyering, just as they do in all other aspects of lawyering.

Unhappy clients file thousands of complaints against their former attorneys every year, and given the importance of settlement, one would expect to see complaints about settlement conduct with some frequency. A careful review of reported cases and of bar complaints, however, paints a very different picture. Complaints against lawyers for substandard settlement advice or behavior are rare. Successful complaints focusing on lawyers’ settlement-related behavior are even rarer. In the ten jurisdictions I studied, there were more than sixteen thousand reported court opinions stemming from cases involving alleged legal malpractice cases over the past decade. Of those, however, fewer than 1 percent related in any way to allegations of settlement malpractice. Furthermore, only a small fraction of this already small set of reported cases included a verdict providing some form of recovery for the former client. The landscape is similar with respect to formal sanctions. Over the last decade, in the jurisdictions I studied, less than 1.5 percent of the almost eight thousand disciplinary actions against lawyers stemmed from the lawyers’ settlement advice or conduct. Even state ethics advisory opinions are similarly silent about this important aspect of lawyering. In short, although clients complain about virtually every other aspect of lawyers’ conduct, they rarely complain about lawyers’ roles in settlement.

This Article explains and casts a critical perspective on that gap, proceeding in three parts. In Part I, I provide support for two empirical assertions: that clients rarely complain about settlement malpractice, and that even when clients do complain about their lawyers’ conduct with respect to settlement, clients rarely prevail.

In Part II, I argue that the current legal malpractice system has three basic assumptions built into its structure, each of which is animated by the limited image of lawyer-as-litigator:

  1. The nature of lawyers’ work demands that they enjoy sweeping deference when they exercise their professional judgment during litigation;
  2. The impact of any lawyer errors will be reflected in changes to trial outcomes; and
  3. Clients’ interests (and, therefore, compensable damages) are reflected in the pleadings and bounded by the court’s remedial powers.

Each of these three assumptions is perhaps sensible if the malpractice system dealt only with alleged litigation errors. Courts will reject, for example, assertions that a lawyer should have called this witness before that witness at trial or should have advanced this theory instead of that theory. Courts will engage in a case-within-a-case inquiry to establish whether a litigator’s decision proximately caused injury, compensating clients only if the lawyer’s conduct changed the outcome of the litigation. And no client will stand to receive compensation for nebulous, speculative, or nonquantifiable injuries stemming from his lawyer’s allegedly inadequate conduct at trial.

The challenge is that these same three familiar—even defensible—principles shield lawyers from virtually all accountability in the context of settlement, rather than litigation. A malpractice system that may work in the lawyer-as-litigator context produces questionable outcomes within the lawyer-as-settlor context. Whether the legal system is designed intentionally to provide unwarranted protection to lawyers is not the animating question of my analysis.7 The current patchwork of rules does, however, serve to make it astonishingly (some might say indefensibly) difficult for any client to prevail with complaints about his lawyer’s settlement-related conduct.

In Part III, I take up the normative question of what a legal malpractice system would look like if it were harmonized with the modern reality of lawyer-as-settlor. Specifically, I argue that a settlement-appropriate conception of lawyers’ roles will require revisiting at least two, and eventually perhaps three, aspects of courts’ treatment of settlement malpractice complaints.

First, courts should cast a critical (rather than deferential) gaze on the settlement counsel lawyers provide. Supporting clients’ autonomy—their ability to make informed choices between different litigation and settlement paths—demands at least three separate tasks of lawyers. Competent lawyers must, in this context, understand their clients’ interests, compare the implications of any proposed settlement against those interests, and articulate the risks and opportunities associated with continued litigation. Each of these may appear obvious, and each finds support in existing articulations of professional ethical obligations.8 As a practical matter, however, courts currently tend to group lawyers’ settlement counseling duties together with their litigation duties, and as a result of the judgmental deference that attaches to litigation decision-making, clients currently find real challenges animating the malpractice system as to even these basic settlement-counseling duties.

Second, a settlement-appropriate malpractice system must recognize that settlements—unlike litigation outcomes—are usually functions of probabilistic rather than binary assessments, are rarely zero-sum, and are ultimately the product of clients’ choices. Furthermore, a case’s settlement value is a function of (but is not identical to) its litigation value. A settlement-appropriate malpractice system would account for each of these differences as it assesses the proper measure of damages, assigning value to the full range of lawyers’ impacts on their clients’ interests.

Third, the current system assigns a level of deference to lawyers’ across-the-table negotiation decision-making that may, at least in some contexts, be unwarranted. The law currently treats settlement negotiation as entirely indeterminate, with every behavioral variation being chalked up to what Justice Anthony Kennedy recently referred to as “personal style.”9 If every conceivable settlement decision can be justified ex post as a matter of style or tactic or context, then lawyers stand effectively immune from any claim that they have breached the duty of care they owe to their clients in this arena. If negotiation is truly indeterminate—if literally anything goes—then it is appropriate that all negotiation should be treated as immune from after-the-fact decision audits. What if it were true that some negotiation decisions make clients demonstrably worse off? Negotiation research and education have made enormous advances in recent decades. I argue that the days of treating settlement negotiations as indeterminate and beyond audit are numbered, if not over.

The reforms I describe here are limited, in the sense that they focus only on the viability of client-driven malpractice actions. Such complaints represent only one component of any professional quality assurance mechanism. They would ideally be accompanied by improvements in education, practice controls, norms, and market forces. But absent some prospect of accountability in instances when a practitioner falls short of appropriate standards, none of the rest of these is likely to be sufficient in assuring profession-wide quality.

The current legal malpractice system makes it harder for clients to bring successful complaints about their lawyers’ settlement conduct than in other lawyering contexts—even though we live in the “age of settlement.10 This luxurious position, for which we have collective responsibility as a self-governed profession,11 should prompt us to talk about the prospect of settlement malpractice openly and honestly. Until settlement malpractice is as rare, in fact, as current litigation trends suggest it to be, the malpractice system should recognize the centrality of settlement to modern lawyering, should assess lawyers’ conduct using settlement-appropriate standards, and should compensate clients for the full injuries they suffer when their lawyers’ settlement conduct falls short.

I. Claims of Settlement Malpractice Are Rare

Lawyers often create value for their clients in the context of settlement. Yet lawyers are human, and it stands to reason that at least some lawyers sometimes fall short in delivering on this promise. Given the volume of settlement activity, one would expect, therefore, to see complaints against lawyers alleging settlement malpractice12 with some frequency in civil claims of legal malpractice, bar disciplinary reports scolding attorneys for their settlement conduct, or ethics advisory opinions clarifying lawyers’ duties related to settlement. In fact, as an empirical matter, we see none of these three.

A. Reported Malpractice Cases

There are currently approximately 1.3 million licensed attorneys in the United States, and the domestic legal services industry comprises more than a quarter of a trillion dollars in revenue annually.13 Each year, there are about eight thousand claims or complaints by clients who were unhappy about some aspect of their lawyers’ services.14 And we know that it is common for lawyers to negotiate and to counsel their clients about settlement.

The empirical question, then, becomes whether claims against lawyers reflect how prominent settlement is in lawyers’ practices. The short answer is no. Less than 1 percent of all reported cases about legal malpractice over the last decade stemmed from claims against lawyers for settlement malpractice.15

Working with available online databases, I reviewed thousands of reported cases involving alleged legal malpractice. Where possible, I automated aspects of the search, but for reasons I describe below and in the Appendix, much of this work demanded individual case reviews in order to discern whether a case stemmed from an allegation of settlement malpractice or from something else. I read16 more than one thousand cases as part of the process of reviewing and coding the dataset.

The decision to use reported cases as the overarching dataset for this portion of my research comes with some virtues and some limitations. Using reported cases has the benefit that the data are easy to access, enjoy a high degree of uniformity across jurisdictions, and are familiar to scholarly legal researchers. Furthermore, because published opinions are the most accessible and common form of information about the treatment of legal questions, they provide the backbone for any public information or deterrent effect. The use of reported17 court opinions for this aspect of the research, however, presents several limitations. For example, it does not capture complaints by unhappy clients whose cases terminated in legally unremarkable, and, perhaps, unpublishable ways. Particularly with low dollar claims that are unlikely to present attractive appellate conditions, the simple legal malpractice claim that dies on an initial dispositive motion, or even on summary judgment, is unlikely to appear as a published court opinion. Similarly, these data would not capture incidents in which clients were unhappy but were unaware that there was even the prospect of recovery against their attorney or that their attorney’s conduct was in some way responsible for the unwanted outcome. These data would not reflect cases in which injured parties were unable to secure legal representation and, therefore, abandoned their claims.18 And of course, data based on court opinions will often fail to capture instances in which the parties to a legal malpractice suit resolve the suit through settlement.19
Pre-suit or early settlement is a routine aspect of modern, complex legal malpractice litigation. Reported cases, therefore, capture only a part of the picture. But I can intuit no reason why legal malpractice claims alleging settlement malpractice should resolve earlier or more quietly than other forms of alleged legal malpractice. This may be a limited snapshot, but because it is comparative as against all other forms of alleged malpractice, these data paint an interesting picture.

The combination of these limiting factors means that this search of reported legal malpractice (1) almost certainly understates the frequency of clients’ unhappiness about their lawyers’ settlement-related conduct, and (2) probably overstates the frequency with which clients’ claims are legally meritorious.20 However, no publicly available, searchable databases of complaints yet exist in a form that would permit research without some or all of the limitations above. It made sense, therefore, to at least include this dataset in the effort to understand the current settlement malpractice landscape.

I recorded as “Settlement Malpractice Cases” only those cases that included an allegation of some lawyer misconduct related to settlement.21 The “Hits” figure represents the number of cases that emerged initially from an automated search, many of which ultimately proved to be false positives. Table 1 below summarizes those findings.

State Total Reported Legal Malpractice Cases “Hits” Using Search Terms Settlement Malpractice Cases Percentage of Legal Malpractice Cases
AL 1,144 75 5 0.44%
AR 698 33 1 0.14%
CT 1,923 154 20 1.04%
DE 1,048 98 8 0.76%
IA 854 37 0 0.00%
ME 398 26 4 1.01%
MD 1,213 62 8 0.66%
MI 2,306 157 43 1.86%
OH 4,275 414 28 0.65%
TN 1,443 88 8 0.55%
WI 1,439 66 0 0.00%
Total 16,741 1,210 125 0.75%

For each of the 125 reported opinions in which the underlying claim was actually about settlement malpractice, I also coded the case according to types of claim, and to the extent discernable, according to which party ultimately prevailed. Those results appear in Table 2.

Resolution Number Percentage
Verdict for Defendant Attorney 84 67.2%
Probable Settlement 3 2.4%
Clear Settlement 4 3.2%
Verdict for Plaintiff Client 9 7.2%
Unknown 25 20.0%

These results are directionally consistent with other avail-able data. I am aware of only two empirical studies that have coded specifically for negotiation and settlement behavior in legal malpractice actions. Professors Herbert Kritzer and Neil Vidmar reported that in the Missouri data they reviewed, spanning the years 2005–2014, just under 10 percent of the complaints against lawyers involved “[s]ettlement or negotiation,” and 22 percent of those complaints were ultimately successful.22 The description of the coding associated with these statistics suggests that “settlement or negotiation” would include a wide range of conduct unrelated to settling lawsuits. For example, an allegation that an attorney failed appropriately to negotiate the terms of a commercial lease would presumably be included on their list, as it involves “negotiation.” Furthermore, their data included complaints that had not yet ripened into lawsuits.23 Similar methodology appears in the American Bar Association (ABA) Standing Committee on Lawyers’ Professional Liability report, which indicates that 5.8 percent of claims against lawyers resulted from lawyers’ conduct in the course of “Settlement/Negotiation,” without differentiating the two activities.24 As a result, the final numbers in these studies include far more complaints than I included in the research for this Article. Nevertheless, even though the datasets and underlying methodologies are quite different from the one I employ in this Article, their conclusions are directionally similar: lawsuits (and particularly successful lawsuits) against lawyers for settlement malpractice are extremely rare.

B. Disciplinary Cases

In an effort to understand the fuller landscape of client satisfaction (or dissatisfaction) with lawyers’ settlement conduct, I also examined formal ethics complaints filed against attorneys. Such disciplinary controls form one of the foundational mechanisms by which the profession seeks to assure the quality of the services its members provide.25 Like malpractice actions, ethics complaints arise only ex post and only when dissatisfied clients take the initiative to bring a complaint. They likely understate considerably, therefore, the instances of lawyer misconduct for all of the same reasons as civil malpractice actions. Furthermore, because ethics complaints rarely provide the prospect of significant monetary recovery for clients, one might imagine that there would be even fewer such complaints, compared with civil lawsuits. Still, enough differences exist—in procedural protections, in barriers to bringing complaints, in decision-makers, in timelines, and in reporting—that ethics complaints form a potentially interesting and distinct dataset for understanding the landscape of settlement misconduct.

One significant feature of any state bar’s collection of disciplinary opinions is that opinions (as opposed to the underlying complaints) almost uniformly reflect only instances in which the relevant disciplinary review board has judged that the attorney’s conduct failed to meet some aspect of the relevant ethical standards. Looking at disciplinary opinions alone, therefore, it would not be possible to determine anything about the frequency with which clients’ complaints are successful.26 One hundred percent of published disciplinary reports reflect a successful complaint—rather than serving as evidence that all complaints are judged meritorious.27 Many states also provide aggregated data that include the total number of complaints filed, and from that, one could determine an aggregated “success” rate across all complaints.28 But even more so than with malpractice actions, there is no reliable way to assess the likelihood that a client’s ethics-based complaint will be successful in contexts in which the alleged misconduct relates specifically to settlement behavior.

The data I reviewed included many instances of attorneys having been accused of multiple kinds of misconduct. For example, in In re Disciplinary Proceedings against Lamb,29 the complaint included at least seventy-five different counts of misconduct against a single attorney. These counts included improperly establishing a client trust account, not returning phone calls, and failure to take action on pending litigation. However, included in what the board called a “pervasive pattern of misconduct”30 were allegations that Lamb had told his client “that the defendants had made a settlement offer” which the client instructed Lamb to accept.31 For years thereafter, Lamb informed his client that the settlement was “close to being finalized,”32 but in fact had forged his client’s signature on the back of the settlement check and deposited the money for himself.33 For purposes of this study, in order to be conservative with the estimates, if any of the allegations against the attorney involved settlement conduct, it was treated as though it were a settlement misconduct case. Table 3 below summarizes the findings.

State Total Published Disciplinary Cases “Hits” Using Search Terms Settlement Malpractice Cases Percentage of Disciplinary Cases
AR 400 256 6 1.50%
IN 1,052 22 2 0.19%
IA 822 326 14 1.70%
ME 211 154 4 1.90%
MA 1,622 775 11 0.68%
MI 1,044 227 14 1.34%
OR 422 417 58 13.74%
TN 1,366 148 2 0.15%
TX 662 321 2 0.30%
WV 310 125 4 1.29%
Total 7,911 2,771 117 1.48%

As with tort-based malpractice claims, clients are unlikely ultimately to prevail with their complaints against attorneys. The chart above suggests that over a ten-year period almost eight thousand complaints against lawyers were successful and resulted in published discipline. But there were more than eighty-five thousand bar complaints filed nationwide in 2016 alone.34 This suggests a “success” rate below 5 percent. Therefore, this review of disciplinary actions yields a conclusion similar to that of the review of malpractice actions: allegations of settlement malpractice constitute only a tiny percentage of disciplinary actions, and clients rarely succeed even when they do file complaints.

C. Ethics Advisory Opinions

Finally, I examined formal state bar ethics advisory opinions, looking to see if settlement malpractice appears in the statements aimed at clarifying the appropriate application of state legal ethics rules.35 Such opinions would not be direct evidence of claims against attorneys by former clients. They might, however, be evidence suggesting some ambiguity in the articulation of ethical constraints or in their application to particular fact patterns. And because such opinions are generally the product of inquiries by practicing attorneys to the relevant body within the state bar, their existence or absence might at least give an indication of the extent to which lawyers fear such complaints.

Because state ethics rules change infrequently, this search involved an expanded timeframe. Whenever possible, I searched every published ethics advisory opinion available since the adoption of each state’s most recent ethics rules or codes. In many cases, this meant that the database extended back several decades. I focused on six geographically and demographically diverse states, with an eye toward including representation both of Model Rule and Model Code jurisdictions.36 In addition to the Boolean search terms used with the other two datasets, I added search terms aimed at unearthing opinions about the ethics provisions that I judged most likely to be relevant to settlement malpractice complaints.37 The results of this search and review process appear in Table 4.

State Years Total Published Advisory Opinions “Hits” Using Search Terms Opinions Involving Settlement Malpractice Percentage of Total Opinions
NY 2008–2017 322 158 4 1.24%
CA 2008–2017 22 14 3 13.64%
IL 2008–2017 63 40 0 0.00%
AL 1981–2017 145 0 0 0.00%
IA 1966–2017 937 8 0 0.00%
DE 1979–2017 79 10 0 0.00%
Total   1,568 230 7 0.45%

Although not identical to the percentage of published court opinions or published disciplinary opinions, it is clear that the data on ethics advisory opinions are not directionally different from those appearing elsewhere in this Part.

A number of factors may contribute to the scarcity of settlement-related complaints. Perhaps clients are unaware that they were injured in any way.38 (“The case settled. I’m happy enough, I guess. You never get everything you want.”) Perhaps clients are unaware that their injuries were the result of their lawyers’ conduct. (“The case didn’t settle because the other side was being unreasonable. What can you do?”) Perhaps cognitive dissonance dissuades clients from viewing their lawyers’ conduct critically.39 (“I make smart choices. I both chose this lawyer and made the final decision about whether to settle. I believe in my choices, and in all events, I made the choices.”) Perhaps clients perceive little prospect of recovery even if they do complain. (“What lawyer is going to help me sue another lawyer?40 What judge is going to rule in favor of a client over a lawyer? And even if I really show some error, am I really going to recover anything?”) Perhaps lawyers are just atypically excellent at all aspects of the settlement process—even though we see lawyers make mistakes in all other aspects of their work, maybe there is just nothing to complain about with respect to their settlement conduct.41

But one would expect that at least sometimes, clients would perceive their lawyers to have engaged in substandard behavior that caused the client some form of injury. To the extent civil lawsuits alleging settlement malpractice are rare even in those contexts,42 and to the extent such suits are unlikely to be successful, one justification might be that although the relevant tort and contract laws do not functionally permit civil liability, the profession’s internal ethical and regulatory mechanisms address any risk of this form of malpractice. To the extent that there are not disciplinary actions against attorneys for violations of their settlement-related duties,43 one explanation might be that the rules are too ambiguous to support complaints reliably. But if it were a question of ambiguity, then one might expect to see a greater number of ethics advisory opinions.44 Instead, the data show that neither complaints nor inquiries are common, suggesting that settlement malpractice is largely not contemplated within the existing legal frameworks.

The scarcity of settlement-related complaints, combined with their futility, paints a potentially troubling picture. If settlement is an important part of modern lawyering, one should expect lawyers to be accountable in those instances when their conduct falls below reasonable professional standards. Given the prominence of settlement to the modern litigation landscape, one should expect the legal system to have developed mechanisms for effectively assessing lawyers’ settlement services for their clients. A critical examination of the current professional liability landscape suggests this is not the case.

II. The (Inadequate) Lawyer-as-Litigator Lens on Settlement Malpractice

The current legal malpractice system relies on a set of at least three implicit assumptions about lawyers, their roles, and the impacts of their decisions. Each of those assumptions is important and even legitimate—if lawyers are acting as litigators.45 First, the current system assumes that lawyers’ judgments should generally be treated with sweeping judgmental deference approaching immunity. Second, the current system assumes that lawyers’ actions will directly—even exclusively—affect their clients’ pros-pects in court. A good lawyer will improve the client’s litigation expectations, and any misconduct will be reflected in a diminished litigation expectation. And third, the current system assumes that the proper measure of any damages deriving from malpractice will be properly captured by comparing the actual litigation outcome with the litigation outcome that would have resulted in a world in which the attorney had acted differently. Each of these three assumptions may be valid and important in a litigation context. But I argue in this Part that each of these assumptions presents serious shortcomings in a world in which settlement is prominent.

A. Breaches of Duty and Judgmental Deference

It is axiomatic that lawyers owe their clients “ordinary care.”46 The litigation context provides a relatively bifurcated categorization of attorney actions. Certain attorney actions are easily measured against a bright-line (low) bar. Missing a statute of limitations, failing to call an expert witness when expert testimony is required, or failing to serve the proper parties will result in a clear finding that the attorney breached her duty of care to her client.47 As to almost everything else, however, courts have treated attorneys’ litigation-related decision-making with a level of judgmental deference that effectively means that the attorney’s decisions are beyond review.48 Litigation is interactively strategic, with each side reacting to the other’s decisions in iterated ways. This dynamic produces a complex system that would be difficult, if not impossible, to unpack reliably from both a duty and a causation perspective. Given this complexity, and because the relevant inquiry is not “was this the best move the attorney could have made,” but rather, “was this an actionably bad move the attorney made,” this level of deference attaches to almost all litigation decisions.

In addressing allegations of settlement malpractice, courts have essentially imported this set of judgmental immunity rules from litigation. The settlement context presents few, if any, bright line rules akin to statutes of limitations. Instead, if anything, the negotiation process is even more strategically interactive than the litigation context. Whether I make this offer or that offer, respond in this way or that way, agree or disagree with this procedural or substantive request, depends in large measure on my assessment of what you are thinking. And meanwhile, you are engaged in the same process of trying to assess me, my reactions, and my decisions. With this image of the challenge of unpacking settlement negotiation decisions, it is not surprising that courts have landed in a place of sweeping deference.

Amplifying this default level of judgmental deference, a lawyer enjoys at least three avenues to defend against an allegation that she breached her duty of ordinary care. First, she may argue that her conduct was common among her peers—that it was, in fact “ordinary.”49 “Everyone is doing it” may not be persuasive as a theory of criminal defense or as an entreaty from a teenager to his parents, but in the context of professional malpractice, the argument can be dispositive because the burden for a complaining party is to establish that his lawyer’s conduct fell outside of normal practice. How can the client demonstrate that this behavior deviates from what other lawyers would have done in that situation?50 These burdens are insurmountable for many complaining parties. Second, a defending attorney may argue that her conduct represented a strategic or tactical decision based on her own experience of what works best for her. “This may not be how everyone else does it, but it’s how I do it, and I know myself and what works for me.” Justice Kennedy’s assertion in Missouri v Frye51 that negotiations are a matter of “personal style”52 has intuitive appeal53 and some empirical support.54 What works for one person in negotiation may not work as well for another. According to this line of argumentation, what may appear to be a breach of a duty might in fact be a highly responsible, adapted, learned set of behaviors that the defending attorney believed to be more effective.55 Third, a defending attorney may point to the uncertainties associated with negotiation and may argue that her decisions were justified from a risk-reward perspective, even if they did not produce the desired results in this particular case. “The actions I took may have been risky, but the potential payoff justified the risk and should not be second-guessed.”56 Under this logic, we should assign judgmental immunity to a broad range of decisions, so long as they are potentially justifiable under some theory of risk and reward, and there is considerable support for this in the current legal landscape.57

Furthermore, because the line between the roles of client and attorney blur in negotiation—in a way that is different from litigation—lawyers often enjoy additional shields from accountability. Consider, for example, the question of what settlement figure to propose to the other side. One can easily imagine that a client might want his attorney to make the strategic decision about when and what to offer, thinking that the attorney’s experience and training and distance from the case better positions her to make that kind of judgment. But one can also imagine a client having a strongly held view about the wisdom of one negotiation approach over another. What then? In the case underlying Ram v Cooper,58 the plaintiffs’ daughters were injured in an automobile accident.59 Unbeknownst to the plaintiffs, their attorney sent a demand letter to the relevant insurance carrier. When the plaintiffs learned of the offer and of the insurance company’s “very discouraging” response, they sought to engage actively in the substance and strategy of the settlement negotiations. Pointing to a practice guide they had consulted regarding typical settlement values and anecdotal evidence of similar claims, the plaintiffs asked their attorney to make an offer five times the daughters’ combined medical expenses—a figure much higher than the figure under discussion between their attorney and the insurer.60 The attorney treated this instruction from his clients as a constructive discharge and withdrew from the representation. The plaintiffs eventually settled with the insurer and filed a legal malpractice action against their former attorney.61 The court found for the attorney, noting that the attorney was not bound to convey an offer he did not believe to be wise.62

In professional activities involving bright-line, scientific rules, there may be unarguably obvious errors. The filing deadline was missed, the architect files a certificate of work that was not actually completed, or the patient’s test results were never appropriately forwarded to a specialist. Many aspects of negotiation, however, have been characterized (accurately) as having at least some aspects that are more like “art” than “science.”63 Negotiation involves iterated decisions with uncertain information, a human client, and a human counterpart, and that combination produces decision points that lend themselves poorly to formulas or bright lines. Whether to make an early offer or wait for the other side to act, for example, may be the subject of strongly held opinions among different practitioners. But no consensus exists among negotiation professionals about a universal prescription regarding whether to be the first to act.64

Reflective of the art aspect of negotiation, some have analogized negotiation to jazz.65 Jazz, like negotiation, relies on some measure of intuition and improvisation,66 paired with experience and practice. Still, it is not the case that there is no such thing as a wrong note or missed timing when one endeavors to play jazz. Some things are honking bad. But as long as the music is at least recognizable as jazz (even if, perhaps, a middle school jazz band version of jazz), its improvisational decisions and implementation will enjoy a heightened level of deference. It is still jazz. Continuing with the analogy, even if there may not be “one right way” to negotiate, it does not follow that there are “no wrong ways” to negotiate. But because of the improvisation and intuition involved, clients face enormous challenges in establishing that some aspect of a lawyer’s settlement negotiation conduct fell below ordinary care.

B. Proximate Causation and Settlement

Even if an unhappy client establishes that an attorney’s settlement conduct fell below the applicable standard of care, he will only be able to recover if he can also demonstrate that the breach was a substantial factor in causing injury to the client.67 The accepted approach for establishing such causation in a litigation malpractice context is the case-within-a-case method. That turn-back-the-clock68 exercise requires an inquiry by a new factfinder, often in a new venue, into what “would have” or “should have” happened in the underlying litigation if the attorney had behaved differently.69 In this inquiry, we hold the rest of the litigation universe constant, testing the effects of a single, new variable—an improved, now minimally competent, level of representation by the client’s attorney. If minimally competent lawyering would have changed the substantive outcome of the litigation, we deem the proximate causation question to have been answered.

In a settlement context, however, there is almost always at least one decision made by someone other than the defending attorney that complicates the question of causation. Three different kinds of actors may make decisions following those of the attorney, and any one of those decisions might complicate or break the causation chain in ways that create sweeping immunity for lawyers.

First, and perhaps most commonly, the intervening decision-maker might be the client himself. Because the final decision whether to settle belongs to the client,70 if a client’s lawyer consents to a settlement, the law generally assumes that the client must have approved it.71 If the other side extended an offer and there was no settlement on those terms, the law assumes that the client must have decided not to approve it. The client, then, must prevail with the challenging argument, “I would have had a more attractive decision to make or would have made a different decision, if my attorney had not mishandled things before the moment when I made my decision.” At least one jurisdiction bars this line of argumentation entirely. As the Pennsylvania Supreme Court wrote in Muhammad v Strassburger, McKenna, Messer, Shilobod and Gutnick,72 “a suit [cannot] be filed by a dissatisfied plaintiff against his attorney following a settlement to which that plaintiff agreed, unless the plaintiff can show he was fraudulently induced to settle the original action.”73 This means that even if an attorney’s pre-settlement actions caused unarguable injury to the merits of the client’s case, decreasing both the expected litigation value and the settlement value of the case,74 a subsequent settlement will cause the client to forfeit any claim against his attorney unless the client can show that his attorney committed fraud against her client in addition to whatever earlier malpractice the client has perceived. Most jurisdictions have declined to follow the bright-line rule in Muhammad and will at least permit clients who have settled to bring malpractice actions against attorneys “if the client can establish that the settlement agreement was the product of the attorney’s negligence.”75 However, a client in these states still has the challenge of demonstrating that he does not now, in hindsight, merely regret the decision to make the deal.76

Second, the intervening decision-maker might be the other party in the lawsuit. An unhappy client might say, “Yes, there was a settlement, but if my attorney had been competent, I would have gotten an even better deal” or “No, there was no settlement, but if my attorney had been competent, the other side and I would have reached agreement.” Both of these assertions rely on demonstrating how the other party would have behaved in an alternate universe in which the attorney’s actions fell somewhere within the broad sphere of ordinary care.

If the client approved—and now regrets—a particular settlement, the client has a high burden in demonstrating that the other side would have ever agreed to something even more favorable, even if the attorney had acted appropriately. For example, in a malpractice action against the attorney who represented him in a divorce proceeding, the client in Marshak v Ballesteros77  alleged that the deal negotiated by his lawyer undervalued the marital residence and overestimated the accounts receivable in his business, in ways that prejudiced him in the settlement.78 Upholding the original settlement and dismissing the legal malpractice action, the court noted that “[e]ven if [the client] were able to prove” that his case was “worth more” than the settlement, “he would not prevail. For he must also prove that his ex-wife would have settled for less than she did, or that, following trial, a judge would have entered judgment more favorable than that to which he stipulated.”79

In a typical lawsuit, how does a plaintiff demonstrate what the other side would have done in a theoretical, alternative settlement context? Absent concrete proposals, as one might see in an offer of judgment context, it is difficult to imagine how a plaintiff could demonstrate what offers would have been made or accepted.80 In the lawsuit underlying Rogers v Zanetti,81 the founders of a home-healthcare company accused James Rogers, one of the company’s investors, of fraud and conspiracy.82 The founders made an offer to settle the suit in exchange for a payment from Rogers of $450,000 and full control of the company, but Rogers’s attorney never communicated the offer to Rogers.83 At trial, a jury found Rogers liable for fraud, and he subsequently sued his attorney for malpractice. Rogers testified that if he had known about the founders’ offer, he “would have tried to settle the case” and “would have instructed [his] attorneys to negotiate the best possible resolution.”84 Rogers’s failure to testify that he would have outright accepted that offer proved fatal.85 Noting the absence of record evidence that Rogers would have accepted the $450,000 offer or that the founders would have accepted a different offer, the Texas Supreme Court found that Rogers could not establish causation.86 Even if a client is legally permitted to present evidence about their theory of what would have happened in a but-for settlement world, the client will face practical challenges in accessing the relevant evidence. As one court noted, “Absent some compelling circumstances, the settling adversary in the underlying case is not likely to admit that, had the lawyer held out, it would have offered substantially more in settlement than was, in fact, offered.”87

Third, in some contexts, someone other than the litigants and their attorneys takes action that implicates the chain of causation.88 For example, courts routinely review the substance of proposed settlements in contexts in which nonparties’ interests may be affected by the terms of the settlement.89 A court’s approval of a settlement does not generally create formal estoppel in a subsequent malpractice action.90 The need for judicial approval, however, does create a proximate causation problem because the client will need to demonstrate that the court which approved one settlement agreement would have approved some other, theoretical settlement agreement. For example, in the case underlying First National Bank of LaGrange v Lowrey,91 the mother of a child and the bank serving as guardian of the child’s estate brought a medical malpractice action against a hospital for injuries suffered at birth.92 The hospital made a settlement offer of $1 million, but the mother’s attorney rejected it before informing her client of its existence. The case went to trial, and the jury returned a verdict for the hospital.93 The mother then brought a legal malpractice action against her former attorney. In order to prevail, she needed not only to prove (1) that the lawyer breached his duty by failing to communicate the offer, and (2) that she would have accepted it if she had known about it in a timely way, but also (3) that the court in the medical malpractice case would have approved that offer, had it been presented as a proposed resolution of the case.94

In short, although the case-within-a-case method of assessing causation may work in a litigation malpractice case, the fit is more challenging with respect to alleged settlement malpractice. The combination of strategic interactivity and the existence of multiple iterated decision-makers complicates the landscape, and the effective result is to make it harder for complaining clients to prevail.

C. Damages and Settlement Speculation

Legal malpractice jurisprudence in the litigation context makes the unarticulated but plain assumption that any harm caused by an attorney’s mistakes will be reflected in harm to the client’s litigation outcome. One compares the actual outcome with the presumed outcome of the case-within-a-case trial, and the difference is deemed the proper damages for the prevailing plaintiff. For example, in the simplest of cases, assume the plaintiff had a slam-dunk claim in an action in which a liquidated damages clause would net the plaintiff exactly $1 million. The plaintiff’s attorney inexplicably failed to file the complaint, and as a result, the claim became time-barred. Assuming the plaintiff can demonstrate through a case-within-a-case trial that his claim would clearly have been successful but for the attorney’s misconduct, the plaintiff would be entitled to $1 million in damages.

In all but the rarest of cases, however, this basic structure does not lend itself well to alleged malpractice in the settlement context. The challenge for courts (and therefore for prospective plaintiffs) becomes one of establishing the damage caused by an attorney’s substandard negotiation decisions. Without an established legal procedure for assessing the negotiation-within-a-case, or some reliable mechanism for establishing what the results would have been in a negotiation that never occurred, courts often deem plaintiffs’ claims too speculative in settlement contexts, even when the attorney clearly engaged in harmful substandard conduct.

In peculiar circumstances, clients may be able to point to a specific measure of damages associated with misconduct. If a client instructs an attorney to accept the other side’s offer, and for some reason the attorney fails to do so, the client should stand to recover exactly the benefit of the foregone offer.95 Or perhaps a client may be able to establish a specific value for at least some part of the injuries he believes he has suffered because of the attorney’s settlement conduct. For example, in Kliger-Weiss Infosystems v Ruskin Moscou Faltischek,96 the plaintiffs believed (incorrectly) that their attorneys had followed their instructions to incorporate an evergreen provision into the terms of a settlement between Klinger-Weiss Infosystems (KWI) and one of its business partners.97 Four years later, when a dispute arose about the eventual termination of the business relationship, KWI was forced to defend its position in arbitration. In their ensuing legal malpractice complaint against their former attorneys, KWI alleged that “but for the defendant’s negligent advice, it would not have executed the [ ] settlement agreement and/or would not have incurred the legal expense of defending the arbitration.”98 The first half of this assertion raises considerable challenges from a damages perspective, because KWI would be forced to demonstrate what the settlement or litigation result would have been in a but-for world. But at least the second half—the expenses of defending a claim in an arbitration that has already occurred—is a knowable, concrete dollar figure that poses no challenge from a damages perspective.

Typically, however, plaintiffs face a real challenge establishing precise damages in an action in which they are asserting that a favorable settlement would have been the result in a world without their lawyers’ misconduct.99 What kinds of evidence, beyond self-serving ex post testimony by the plaintiff, would satisfy the burden of demonstrating that there would have been a different offer and that it would have been accepted? As the court in Filbin v Fitzgerald100 noted, “The requirement that a plaintiff need prove damages to ‘a legal certainty’ is difficult to meet” because “[e]ven skillful and experienced negotiators do not know whether they received the maximum settlement or paid out the minimum acceptable.”101

Rather than wade through speculation about the precise back-and-forth that would have occurred in a negotiation that never occurred, courts may sometimes permit expert testimony about what cases “like this” settle for “normally.” Repeat players in litigation provide one context in which such calculus may be appropriate. For example, an explosion in 2005 at BP’s Texas City refinery killed fifteen workers and injured almost two hundred others.102 BP “made the decision to settle every case arising from the plant explosion.”103 Jose Elizondo was among the injured workers, and his attorney negotiated a $50,000 settlement. Elizondo subsequently alleged that the settlement was “inadequate” and was the product of having been “sold down the river” by his former attorney.104 To assess the merits of Elizondo’s assertion that the settlement was insufficient, the Texas Supreme Court opened the door to probabilistic expert testimony in the legal malpractice action, writing:

Here, where the same defendant settled thousands of cases, and indeed made the business decision to settle all cases and not try any to a verdict, we see no reason why an expert cannot base his opinion of malpractice damages on a comparison of what similarly situated plaintiffs obtained from the same defendant. This data is perhaps the best evidence of the real-world settlement value of the case.105

Outside of the context of legal malpractice claims, a small but growing industry has developed around the question of the “value” of pending litigation or prospective cases.106 The initial impetus for such ventures centered on litigation finance, with the prospect of more efficiently allocating investment resources against litigation representing desirable risk profiles.107 In many cases, the algorithms underneath these assessments have proven more reliable than the judgment of even experienced attorneys.108 Free, online versions of case assessment algorithms now exist. Even the most commonly used legal research engines now provide services that include data about settlements and verdicts.109 I am not aware of any court using such services in the context of a settlement malpractice case, but it is not a leap to imagine that expert testimony (such as that which is provided in breach of contract or intellectual property infringement cases) might form the basis for one means of at least putting parameters around malpractice damages.

Absent exceptional circumstances or persuasive apples-to-apples external benchmarking, however, a plaintiff will have a real challenge in proving that the other party in this case would have made an attractive offer, that this client in this case would have accepted it, and the specific terms of that hypothetical offer.110 The result is that our malpractice system, which contemplates the lawyer-as-litigator, makes it tremendously difficult for any client to bring a successful action against his lawyer for alleged settlement misconduct.

III. The Modern Reality of Lawyer-as-Settlor

The current legal malpractice system falls short with respect to lawyers’ settlement conduct. But it need not be so. Lawyers’ settlement-related duties are clear enough that ex post examination is both possible and appropriate.111 The effects of lawyers’ settlement advice and conduct are not so hopelessly complex to be treated as intolerably speculative. The prescriptions from the negotiation literature are virtually uniform. Applied to the lawyer-as-settlor context, these best practices create conditions in which clients can reasonably expect competent services from their attorneys. And attorneys who fail to adhere to these practices—both with respect to the advice they provide to their clients and with respect to the conduct of settlement negotiations—ought not to enjoy the de facto immunity that attaches to a malpractice system that is based on the vision of lawyer-as-litigator.

A. Lawyer-as-Settlement-Advisor: Helping Clients to Weigh the Prospects of Settlement

Properly understood, a lawyer has three distinct but related roles with respect to her client’s decision-making about settlement. The lawyer must understand her client’s interests—the “needs, desires, concerns, and fears” that motivate negotiators’ decisions,112 akin to what the Model Rules describe as the “objectives of representation.”113 She must communicate the implications of any potential settlement—what the negotiation literature refers to as “options”114 —with a specific eye toward the ways in which that settlement option intersects with the client’s interests. And she must provide the client with a candid and accurate picture of litigation’s risks, costs, and opportunities.115 A lawyer’s mishandling of any of these represents—or at least should represent—both a breach of the lawyer’s ethical duties116 and a clear instance of professional malpractice.117

1. Interests.

At the core of an attorney’s ability to advise her client effectively is an understanding of the client’s interests. Because settlements can—and often do—have terms that exceed the court’s narrow remedial powers, clients face a complex decision. On the one hand, litigation presents some probability that the court will exercise its power in a way that redistributes property or rights between the litigants. On the other hand, a potential settlement represents a consensual allocation of property or rights between the litigants, potentially in ways that do not align precisely with how a court might allocate them. Making the choice between these two possible paths raises apples-and-oranges problems. The common currency, the means of comparison from the client’s perspective, is the degree to which each satisfies his most important interests.118 In order to serve a useful settlement advisor function, therefore, a lawyer must understand those interests.

2. Settlement options.

In order to exercise informed choice, a client must understand the implications of a proposed settlement option, and in particular examine the ways in which it does or does not satisfy the client’s interests. A number of dollars, a particular release or waiver clause, a confidentiality provision, a dispute resolution clause, a joint press statement, a contingent clause—any might be options that would take effect only if both parties agreed to them, and each might affect the client’s interests. Assessing and articulating the implications of options, therefore, becomes central to the role of lawyer-as-settlement-advisor.119

At a bare minimum, an attorney must convey the basic terms of an agreement accurately to the client. In the case underlying Arnav Industries, Inc Retirement Trust v Brown, Raysman, Millstein, Felder & Steiner,120 for example, the terms of a settlement were reached, but the next day, the plaintiffs’ attorney sent a revised agreement for their signature, indicating that there had been a typographical error in the original.121 The plaintiffs signed the revised agreement without rereading it, only later to discover that the sheets they signed included a number of changes resulting in a roughly $4 million reduction to what the plaintiffs would be owed upon the default of the opposing party to the settlement.122 The court in the subsequent malpractice lawsuit held—appropriately—that this plainly states a cause of action against the plaintiffs’ original attorney.123 Attorneys must also accurately explain certain conspicuous deal terms, such as an amount of payment124 or the identities of the parties who will be bound.125

More commonly, a client may not understand the implications of some inconspicuous deal term, particularly to the extent it may have an effect on the client’s interests extending beyond the immediate case in question. For example, in the suit underlying Ordon v Karpie,126 Dr. Andrew Ordon faced a complaint before the Connecticut Medical Examining Board.127 His attorney negotiated a potential settlement according to which Ordon would pay a $2,500 fine to the State of Connecticut but would face no restrictions on his practice.128 Indeed, Ordon’s attorney “told Dr. Ordon that accepting the [o]ffer would have ‘essentially no import’ on him or his practice in Connecticut or any other jurisdiction.”129 Relying on this advice, Ordon accepted and paid the fine. However, on the basis of the settlement’s consent order, both California and New York subsequently instituted reciprocal discipline on Ordon that impaired his ability to establish practices in those states. Ordon alleged that he would not have agreed to the settlement in Connecticut if he had known about the implications elsewhere.130

Similarly, in Collas v Garnick,131 the plaintiff was injured in an automobile accident and brought a claim against the owner of the other vehicle involved.132 The plaintiff’s attorney negotiated the terms of a potential settlement and presented it to his client, assuring her that “the release would have [no] impact upon her plan to sue the manufacturer of the vehicle in which she had been riding.”133 The plaintiff signed the settlement and subsequently brought an action against the manufacturer of the seat belts in her car. The trial court barred that subsequent action on the basis of the general waiver she had already signed. She, therefore, brought a malpractice action against her attorney for breaching his duty to provide her with accurate counsel about the extent to which the proposed settlement would satisfy her interests.134

The scope of a lawyer’s advisory duties in this context is not, in most contexts, so abstract and boundless as some courts appear to suggest. Instead, the proper question is simply, “What impacts would this proposed settlement package have on my client’s interests, as I understand them?”135 No lawyer has discharged her lawyer-as-settlement-advisor duties professionally if she has not provided information about how a proposed settlement would address the client’s relevant interests.

3. Nonsettlement alternatives.

In order to decide wisely about whether to settle a lawsuit, clients must have an understanding of their alternative(s) to settlement. Each party is presumed to have a number of different courses of action it could conceivably pursue if there were no agreement—courses of action that do not require the consent of the other side. Proceed with litigation as planned? Call a press conference? Join other parties to the litigation? Add claims or defenses? Forfeit the litigation? From among this list, whatever specific course of action a party deems to be best at satisfying its interests is deemed that party’s Best Alternative to a Negotiated Agreement (or BATNA, in the nearly universal parlance of the literature).136

If a lawyer provides poor advice about the client’s BATNA, the client is plainly deprived of the ability to make an informed choice. A lawyer tells her client, “I’m ninety-nine percent sure this case is going to be won” and “[t]here’s always that one in a million chance that some fluke strange thing can happen and you’d lose but it’s not going to happen here,”137 but then the client loses the case. Another lawyer advises her client that a settlement is reasonable based on expected litigation outcomes—even though she had not investigated the defendant’s available assets.138 A client followed a third lawyer’s advice and settled with three defendant landlords for a total of $2,500 in a lead paint case—even though subsequent investigation revealed that the attorney had never even served one of the parties and that the unserved party carried a $300,000 insurance policy to cover such claims.139 A fourth attorney urges a client to resist paying their neighbor’s request for $19,000 to contribute to repairs on a shared, private road. The client winds up owing the neighbors the full amount of the repairs, owing their own attorney $380,000, and owing the neighbors’ attorneys roughly $580,000 under a fee-shifting arrangement.140

The best practice for attorneys in helping their clients to exercise informed choice in the face of the future uncertainties inherent in litigation likely involves the use of some form of litigation risk analysis (LRA). LRA methodology builds on decades of research and application in a wide range of fields beyond litigation.141 Although LRA can be completed effectively in its basic forms with no more than middle school algebra, specialized software has existed for at least two decades to assist lawyers and other professionals in precisely this kind of activity.142 LRA can serve a “clarifying and calibrating”143 function by attaching numbers to a lawyer’s adjectival pronouncements. It can provide the kind of visualization without which “a client may simply be un-able to process the available information readily due to the degree of complexity.”144 And it holds the prospect of helping lawyers and clients to understand the impacts of probability distributions, unearthing situations in which clients are sensitive to particular outcomes or ranges of outcomes.

There are, of course, complications with litigation risk analysis. Attaching numbers risks creating a false sense of certainty, particularly if LRA is used to summarize a case into a single expected value. A client who is told simply that his case is “worth $1,500,000” has no sense of the shape of the distribution curve leading to this probabilistic conclusion, and even no sense of the likelihood of a $0 recovery or some other extreme outcome. A heavily quantitative approach risks inaccurately suggesting that financial payoffs are the only (or even most important) interest a client has, or should have, in deciding on a proper course of action. Virtually all clients have interests that extend beyond monetary terms. Even if it is possible to reduce nonmonetary interests into monetary valuations (“achieving certainty this year is worth $100,000 to me” or “not setting a bad precedent is worth paying double on this settlement”), the reduction to a single monetary term risks suggesting false or misleading equivalents for clients.

Furthermore, even if one is confident in the basic notion of LRA, lawyers’ ability to conduct such analyses on behalf of their clients is imperfect. If the variables on which the formulas rely are unreliable, the products of the underlying calculations will also be unreliable. “Garbage in, garbage out” may overstate the matter, but not entirely. Lawyers face a number of structural and analytic barriers to producing reliably accurate predictions,145 and these challenges have been demonstrated empirically.146 In some contexts, experienced lawyers may be able to dampen some of the biases that tarnish these predictions.147 And there is reason to believe that the advent of big data analytics will improve lawyers’ ability to make more accurate litigation predictions.148 Still, even experienced lawyers, armed with identical information about cases, commonly produced widely varied predictions about legal outcomes. In short, many lawyers may be quite inept at providing the kind of predictive services on which LRA depends.

From a malpractice perspective, there are complications about the degree of accuracy we can reasonably expect. What buffer or margin of error is tolerably within the scope of “reasonable care” when making such predictions?149 If a lawyer tells her client that he has a 10 percent chance of recovery, and a panel of experts subsequently determines that the client’s chances of recovery were actually 15 percent, one would expect that we would not consider that lawyer’s assessment to have been negligently mistaken. One would expect that if the chances of recovery were actually 90 percent, the client would be justifiably disappointed in the lawyer’s advice. The law has no clear line about the precision and accuracy clients ought to be able to expect of their attorneys’ predictions.

The fact that LRA presents complications, however, does not lead to the conclusion that clients cannot reasonably expect some degree of specificity or accuracy in their lawyers’ predictions. For clients to exercise meaningful choice, they must have some means by which to compare alternative paths. Unlike litigation decisions, which are plainly strategically interactive in a way that merits some level of judgmental deference, attorneys’ advice to their clients holds no strategic interactivity. Such advice, therefore, does not deserve the same level of deference. And it is so central to the role of lawyer-as-settlement-advisor that some set of professional expectations must attach to it.

B. Lawyers, Paper Trails, and Accountability for Oral Advice

Each of the three duties described above (exploring the full range of relevant interests, assessing settlement options against those interests, and articulating the risks and opportunities associated with litigation) is central to the role of lawyer-as-settlement-advisor. Each has grounding in both existing articulations of legal ethics and the negotiation literature. One might imagine that with clear duties, it would be relatively easy to demonstrate breach or compliance.

And yet, my review of the 125 malpractice cases described in Part I suggests that there is often no shared understanding of what the lawyer actually did, much less whether that (alleged) conduct measured up to the relevant standard of care.150 In short, many in the legal profession—particularly in the settlement context—appear to operate routinely without any discernable paper trail or other dispositive record about their conduct or advice.151 The initial fight, therefore, in a settlement malpractice contest, is often not about whether a lawyer’s advice or conduct was adequate, but rather about what the lawyer’s advice or conduct actually was.

Confidentiality is foundational to most conceptions of the lawyer-client relationship.152 It does not follow from that, however, that lawyers’ advice to their clients—or the protocols for developing that advice—necessarily must be unwritten. When I went into the doctor recently for a relatively simple medical procedure, my physician gave me a form outlining her diagnosis, her recommendation, and the expected results and risks associated with the procedure. Roughly the same was true when I recently took wakesurfing lessons. And the same was true when I went in for physical therapy (not unrelated to the aforementioned wakesurfing). When I have received savings advice from my investment advisor, I received her summary in writing, and the same was true when I hired someone to fix the cabinets in my kitchen. On what basis has the legal profession resisted this relatively low-cost mechanism to improve the quality and reliability of this critical information?

Perhaps the process of advising clients about settlement would lend itself well to the development of a practice-informed checklist. Although some professionals initially balk at the notion that their practice has any routine aspects that would lend themselves to such summaries, Atul Gawande recounts in his publication of The Checklist Manifesto that airline pilots have long used pretakeoff checklists, refined through years of testing.153 Checklists have gained prominence in settings ranging from high-end cuisine to rock concert productions.154 Notably, in one set of experiments, physicians across the world saw a 36 percent decrease in major surgical complications and a 47 percent decrease in surgical death rates after the introduction of checklists.155 Such checklists contain nothing novel.156 Indeed, their purpose is not to introduce new ideas, but rather to assure—really assure—that everyone involved is on the same page, particularly when the stakes are high. Why not in settlement conversations?

Whether in checklist form, an opinion letter, or even as part of a routine intake form, it is not difficult to imagine a paper trail version of any of the three duties listed in the subsection immediately above. A client with the opportunity to review how his attorney has articulated or summarized his interests has the chance to correct misunderstandings before they ripen into injury.157 An attorney with a well-written explanation of the implications of a proposed settlement simultaneously serves her client’s interest in informed consent and protects herself from a subsequent accusation that she did not adequately explain the agreement. And much of the point of litigation risk analysis is the creation of a shared visual, which presumes that it is written or otherwise recorded. Even if the profession has not yet developed the norm of conducting formal litigation risk analysis as part of settlement counseling, it is plain that every articulation of lawyers’ duties includes the obligation to assure that the client is sufficiently informed about the choices he faces.

The most common objection to the creation of paper trails is that they involve additional transaction costs—costs clients are unwilling to bear. At the conceptual level, it is unarguably true that there would be some additional cost associated with producing written advice. In a small-stakes piece of litigation, any additional attorney time could constitute a meaningful percentage of the value of the dispute. I am skeptical, however, of this as a blanket justification or explanation. Recall, we are assuming that the attorney has already invested the appropriate hours to formulate competent advice. The question is about the marginal investment of time required to write that advice down—to capture that which the attorney has already communicated orally to the client. Compared with the array of other legal expenses associated with most forms of litigation, a written record of settlement-related advice is unlikely to constitute the kind of factor driving litigation costs for clients.

What if it were the norm for lawyers to produce written records of their settlement advice?158 In most cases, one would expect to see malpractice actions leapfrog over the “my lawyer told me X”—“did not”—“did too” phase, directly to the question of the advice’s adequacy. Lawyers would be protected from inaccurate characterizations of their oral advice, and clients would be protected from lawyers mischaracterizing their oral advice. Perhaps there would be some risk that written accounts of the lawyer’s advice would be at such a level of abstraction that clients would derive no meaningful guidance. (“All litigation comes with risks.”) Or perhaps the paper trail would go the opposite direction, including so much boilerplate, detail, or jargon that clients would similarly find no real guidance.159 One would expect that a client who received unhelpful oral advice of either of these flavors from his attorney would object and seek greater specificity.160 Given the risks of miscommunication or of subsequent disputes about the contents of that advice, it is hard to conceive that lawyers should stand apart from the common practices of so many other professions.

C. Recognizing the Full Value of Settlement Choices

A malpractice system harmonized with the realities of modern settlement lawyering would not only recognize the lawyers’ duties I enumerate in Part III.A, but also would include a recognition that lawyers’ misconduct can create injuries in ways or in degrees different from their impacts in a litigation context. Settlements, unlike litigation, are rarely winner-take-all, and instead are often the product of probabilistic assessments and corresponding allocations. Settlements, unlike litigation, often produce non-zero-sum outcomes. And settlements, unlike litigation, place clients conspicuously in the posture of having agency or choice over the path they wish to pursue. In this Section, I return to each of these aspects of modern settlement, with an eye toward envisioning a malpractice regime consistent with those realities.

1. Accounting for probabilistic assessments.

The most problematic feature of the case-within-a-case approach is that it reduces cases to binary, all-or-nothing outcomes.161 As a result, it misallocates damages in many litigation contexts and is even worse with respect to settlement. The good news, though, is that by acknowledging simultaneously the roles that probabilities and settlement play in modern litigation, both of these problematic features are dampened, if not eliminated.

To illustrate the challenge of the current system, consider four hypothetical cases in which lawyers who engage in roughly similar malpractice are subject to wildly different treatment. In each, assume that the plaintiff will receive $1 million if he wins, and $0 if he loses. Assume that there are no litigation costs, and that we are able to know with certainty the likelihood of each possible litigation outcome.162 The lawyer in our hypothetical commits an error163 that has the effect of decreasing the chances that the plaintiff, her client, will win at trial.

  1. Scenario 1.  Prior to the lawyer’s error, the plaintiff had a 70 percent chance of prevailing, but as a result of the error, the plaintiff entered the trial with a 40 percent chance of winning. We will assume, therefore, that in 60 percent of such cases, the plaintiff loses and brings a malpractice action against his attorney. The current system’s case-within-a-case approach would conclude that the plaintiff in every one of those malpractice cases “would have” won in the but-for world in which the attorney did not make a mistake, because the plaintiff’s original odds of winning were greater than 50 percent.164 The plaintiff, therefore, would stand to recover the full value of the litigation—$1 million.
  2. Scenario 2.  Prior to the lawyer’s error, the plaintiff had a 40 percent chance of prevailing, but as a result of the error, the plaintiff entered the trial with a 10 percent chance of winning. There is a 90 percent chance that the plaintiff will lose the case, and, as in Scenario 1, we will assume that the plaintiff in these cases brings a malpractice action against the attorney. The current system’s case-within-a-case approach would conclude that the plaintiff “would have” lost even in the but-for world in which the attorney did not make a mistake, because the plaintiff’s original odds of winning were less than 50 percent. The plaintiff, therefore, would stand to recover nothing against the attorney—even though the attorney unarguably decreased the plaintiff’s likelihood of winning. The law treats the client here as though he suffered no injury.165

Scenarios 1 and 2 involve identical errors, producing probabilistic harms of identical magnitude. In each, the attorney’s error caused her client’s prospects at trial to drop by 30 percent. But the current system treats the plaintiff in one as though he deserves to have the attorney serve as guarantor of the entire value of the case, while treating the plaintiff in the other case as though he suffered no injury at all. In this sense, Scenario 2 mirrors the scenarios that have given rise to the debate over the “loss of chance” doctrine.166

What the current system is missing is a recognition of the roles that probability and settlement routinely play in the course of modern litigation. There is a distinction between a case’s litigation value (LV) and its settlement value (SV), with the latter being a function of the former. But in a simplified world, one might assume that a decrease in a case’s expected LV would have a corresponding negative impact on its SV, and so one could helpfully consider the effects of the lawyer on her client’s case’s LV.167

The lawyer’s actions in both Scenario 1 and Scenario 2 decreased the LV of her client’s case by $300,000.168 Recovering this amount—rather than the $1 million or $0 result suggested by the case-within-a-case method—would truly return the plaintiff to the position he was in before his lawyer committed malpractice.169

In making this observation, I do not mean to suggest that in the real-world we could assume that the plaintiff would have been able to reach a settlement in the case for any particular amount, much less the specific amount represented by the case’s LV. What I do urge, however, is that a modern legal malpractice system ought to compensate—rather than overcompensate or undercompensate—clients with cases whose value was diminished by the actions of their attorneys. And by incorporating the probabilities involved, along with the prospect of settlement, the system achieves the goal of returning the injured client to the position he was in before the malpractice occurred.

When the attorney misconduct concerns settlement, rather than litigation, the same principles apply, but their application is more complex. For purposes of simple illustration, I focus here on the circumstance in which an attorney provides her client with actionably poor advice about the client’s prospects in litigation.170 As above, incorporating probabilities and settlement dampens the lottery-winner effects of the all-or-nothing case-within-a-case approach built into the current malpractice system.

Assume for purposes of Scenarios 3 and 4 a lawsuit in which the lawyer’s client, the plaintiff, stands to net $1 million if he wins, and $0 if he loses. Assume no transaction costs, and assume further that the defendant has offered to settle the case for $500,000.

  1. Scenario 3.  The lawyer informs the plaintiff that his odds of winning at trial are 60 percent. Based on this advice, the plaintiff client rejects the defendant’s settlement offer. In fact, the lawyer’s assessment was wrong, and the plaintiff actually had a 30 percent chance of winning at trial. The client loses this lawsuit 70 percent of the time, and we will assume that he brings a malpractice action against his attorney. Under the current system’s approach, the client would testify (presumably credibly) that he “would have” accepted the defendant’s offer if he had been properly advised about his BATNA. Comparing that but-for world with the world that actually occurred, in which the plaintiff lost the case, the current system would award the plaintiff the difference—namely, the full amount of the defendant’s offer: $500,000.
  2. Scenario 4.  The lawyer informs the plaintiff that his odds of winning at trial are 30 percent. Based on this advice, the plaintiff accepts the defendant’s settlement offer. In fact, the lawyer’s assessment was wrong, and the plaintiff actually had a 60 percent chance of winning at trial. Under the current system, if the plaintiff is able to identify this error, he will (presumably credibly) testify that he would have rejected the defendant’s offer if he had been properly advised by his attorney. The case-within-a-case approach would then say that in the but-for world the plaintiff “would have” won the underlying case, receiving $1 million. The difference between that but-for outcome and the actual outcome (the plaintiff already holds $500,000 from the settlement) is another $500,000.

What makes more sense—and is more aligned with the realities of modern litigation—would be to put the plaintiff in the same position he would have been in before the lawyer delivered her substandard advice. The injury to the plaintiff is the difference between the actual result the plaintiff obtained and the value of the choice the plaintiff would have made, if properly advised.

In Scenario 3, the plaintiff’s actual choice at the moment before the lawyer’s misconduct was between a $500,000 settlement and going to trial, the LV of which was $300,000. The client would have presumably accepted the settlement. This outcome should then be compared with the value of the endowment the client actually held after his decision based on his lawyer’s faulty advice. In this case, the client still held the value of bringing the case to trial—a LV of $300,000. Before knowing what the result of the trial would be, the plaintiff would have presumably been indifferent between going to trial and settling (or selling the right to go to trial) for $300,000.171 So, the true injury to the client is the difference between what the client could have had ($500,000) and what he was actually left with ($300,000)—namely $200,000.172

In Scenario 4, the same logic produces a similarly dampened result. The client believed that his choice was between (1) settling for $500,000 and (2) proceeding to trial—an endowment he believed worth $300,000. He chose to settle for $500,000. But his actual choice was between settling for $500,000 and proceeding to trial, which had an expected value of $600,000. If the system provides him with the difference between those two values—in other words, $100,000—he will be in the same position he would have been in before the lawyer’s misconduct.

Two immediate challenges arise about the prospect of this kind of assessment. The first is that in the real world, it can be difficult to assess with precision both the payout and the probabilities associated with future litigation.173 Each participant in a piece of litigation has access to different information, different experiences coloring their interpretation of that information, different incentives, and different potential cognitive biases that are likely to cause the exercise to lack scientific precision. Still, clients and attorneys are called upon to make probabilistic assessments, based on imperfect information about expected payouts, every day. To ignore the prominence of those decisions is to ignore a fundamental aspect of modern litigation.

Second, and perhaps more importantly, in an Article broadly aimed at urging greater accountability for lawyers regarding their conduct in settlement contexts, it may seem odd to urge a system in which some injured plaintiffs, in some contexts, would receive less in damages than the current system provides. But the system adjusted for probabilities and settlements does not always produce a smaller recovery for plaintiffs. Furthermore, I might hope that if the malpractice system had fewer lottery-ticket attributes, we might see greater willingness from the courts to examine honestly the full range of impacts from lawyers’ conduct. The conundrum of the current system is the manner in which it treats decisions and outcomes as binary, all-or-nothing. The reality is that litigation involves probabilities and the prospect of settlement, and by incorporating those factors into the system, we better compensate (without overcompensating arbitrarily, or failing to compensate at all) those clients who are injured by their lawyers’ actions.

2. Accounting for value creation.

A case’s litigation value and its settlement value are too often conflated. In some respects, calculating the value associated with a case and its resolution may be relatively straightforward: a case’s settlement value equals the litigation value of a case, plus or minus the anticipated risk-adjusted transaction costs associated with continued litigation.174 This is familiar, indeed foundational, to much of the classic economic analysis of the settlement of litigated cases.175 Transaction costs, fully understood, include both direct, easily quantifiable transaction costs such as attorneys’ fees and less direct opportunity costs such as the time and attention litigation consumes, depriving people of the ability to otherwise be engaged in productive activities. Even before adding in the possibility that disputants will have mismatching predictions about likely litigation outcomes,176 the simple prospect of avoiding transaction costs makes settlement Pareto dominant over litigation in many conditions.177 But in practice, assessing a case’s settlement value includes more factors than this. Indeed, the negotiation literature describes a range of ways in which a client’s valuation of both litigation and settlement extend beyond a simple economic calculation of expected net trial outcomes—even in contexts in which the client is acting wholly rationally.178 Broadly speaking, settlements often produce non-zero-sum outcomes or “value” for the parties that represents a greater total utility than litigation could have produced.179

There is some evidence to suggest that people are not universally good at creating value in negotiation contexts.180 But the fact that some people negotiate in ways that create suboptimal outcomes does not mean that we should expect or condone such behavior—particularly among those for whom this forms a significant part of their professional practice. There are at least some examples of courts holding that attorneys have a professional obligation to help their clients to capture some of these kinds of value-creating deals. For example, a settlement’s structure or timing may have profound, but different, tax consequences for the parties, raising the prospect of efficient trades.181 Lawyers for litigants locked in an intellectual property dispute might help clients to structure a forward-looking business arrangement involving licenses and royalties as a means of capitalizing on non-zero-sum opportunities between the disputing parties.182 Counsel may (correctly) anticipate the prospect of future disputes arising during the implementation of complex settlement terms and may build in predetermined dispute resolution processes to minimize the transaction costs associated with those implementation problems. The settling parties may be able to take advantage of economies of scale, permitting the defendant to provide some benefit to the plaintiff at a lower cost than the plaintiff would need to bear on his own.183 Or the parties may have nonmonetary or nonquantifiable interests such as protecting their reputations.184

In addition to the prospect of value creation, almost every settlement negotiation will also present distributive questions (such as how much money will exchange hands between the parties).185 A lawyer’s negotiation missteps may also harm her client in this aspect of the negotiations. Consider, for example, the case underlying Bonifer v Kullmann Klein & Dionenda.186 Faced with a personal injury claim by an injured shopper, Wal-Mart conveyed an offer of $35,000 to the plaintiff’s attorney, and the attorney indicated that the plaintiff would accept the deal. The plaintiff subsequently expressed frustration with Wal-Mart’s offer and even tried to argue that the attorney had no authority to signal acceptance. But when the plaintiff pressed for more from Wal-Mart, Wal-Mart refused. The plaintiff suspected the attorney’s disclosure of privileged information in negotiations had undermined the settlement efforts.187 Because offers of compromise are generally inadmissible at trial,188 an attorney’s actions in a case like Bonifer would have had no impact on the case’s litigation value. Even in the context of a case that was purely zero-sum, however, it is reasonable to assume that an unauthorized189 disclosure such as this would have a negative impact on the case’s settlement value.

What would it look like for a malpractice system to recognize that settlements are often non-zero-sum and that a case’s settlement value is a function of (but not identical to) its litigation value? At a minimum, clients would be able to recover damages beyond those narrowly falling within the remedial powers of the court. The case-within-a-case analysis may serve properly to unearth and quantify the impacts of lawyers’ litigation decisions. But in a non-zero-sum world, clients’ recovery should not be limited to that which a court might have done. Rather it should be a function of the utility the client would have received from the now foregone settlement.190

3. Accounting for client choice.

In civil litigation—or in a civil malpractice action stemming from that litigation—does client choice itself currently have a legally cognizable value,191 separate from injury to the litigation value or settlement value of a case? Under the current jurisprudence of legal malpractice, the answer is plainly no. For example, in Moores v Greenberg,192 Moores hired a contingent fee attorney to help him bring a claim for injuries he sustained on the job.193 His attorney received a $90,000 settlement offer from the defendants, but never communicated the offer to Moores, later saying, “the sums mentioned to him were too niggardly to be relayed.”194 The case went to trial, and the jury found for the defendant. Moores later asserted that “he would have accepted the $90,000 offer had he been informed of it.”195 To what damages should Moores have been entitled, if he could prove the other elements of his malpractice claim? The court held that the attorney, “by failing to communicate the offer . . . effectively deprive[d] his client of the net benefit of the tendered bargain—nothing more.”196 Even if he were to succeed with all of his claims, therefore, Moores could receive at most $60,000—the amount Moores would have netted from such a settlement, once reduced by the lawyer’s contingent fees. The First Circuit spent considerable time detailing, and ultimately rejecting, Moores’s argument that his award in a legal malpractice case should not be reduced by the fees he allegedly owed to the very attorney who committed malpractice.197 But nowhere did the court appear even to consider the possibility that the lawyer’s action, by depriving Moores of the ability to make a choice, had caused some separate injury.

The law is not otherwise ambiguous on the abstract question of whether lawyers should help (rather than hinder) their clients’ ability to choose. “[T]he objectives of representation” are specifically the province of the client, according to virtually every articulation of legal ethics,198 and lawyers are expected to “explain a matter to the extent reasonably necessary to permit the client to make informed decisions regarding the representation.”199 Courts have recognized, as a matter of agency, the importance of assuring that clients retain the ability to make independent choices about settlement.200 And courts have refused to enforce provisions in contingent fee arrangements that would encroach on a client’s ability to exercise choice.201 What courts have not done is move beyond a damages calculation that compares (1) the substance of the outcome in which the client had no adequate choice with (2) the substance of the outcome that would have resulted from the client exercising choice. In other words, courts provide little support in practice for the notion that a lawyer who deprives her client of the ability to make a choice creates an injury independently worthy of remedy.202 This result is inconsistent with the modern reality of settlement—a condition in which clients routinely ask lawyers to help them make good settlement decisions (but not to strip them of those decisions).

D. Lawyer-as-Negotiator: Beyond Indeterminacy

As I described above in Part II.A, lawyers’ settlement decisions are often treated with almost complete judgmental deference. And in Part III.A, I argued that at least many of the decisions lawyers make with respect to settlement counseling (the behind-the-table aspect of settlement negotiations) do not merit this level of deference. I left aside the harder question of whether there are across-the-table settlement negotiation actions a lawyer may take that would be subject to less sweepingly deferential scrutiny. Might attorneys-as-negotiators ever face greater scrutiny than they do today?

Negotiation’s strategically interactive nature creates limits on the degree to which we can effectively assess behavior. A professional system might employ ethical or other constraints to create boundaries around a negotiator’s acceptable behavior. But so long as multiple choices are available, and no equilibrium or payoff structure exists to make one of the choices demonstrably worse, then no negotiator’s decision can be subject to ex post critique as irrational or substandard. Whether a particular move was “right” depends on what the other side did, which in turn depends on what they thought you would do. Negotiations between human beings, therefore, will never lend themselves to a formulaic, paint-by-number approach in which specific moves are always prescribed.

But it does not follow from this that there is no hope for improvement or for articulating standards below which competent professionals must not fall. Indeed, just because we cannot perhaps name an ideal approach to negotiation does not mean that one cannot identify practices that fall below the level of professional competence. In this Section, I name several reasons I am hopeful that the landscape will change in directions that will provide higher quality settlement-related services for clients.

1. Potential lessons from specialized civil practice.

Most of the analysis in this Article has focused on the rules and standards applicable to all lawyers, in all civil settlement contexts. These are not the only standards governing lawyers, however. Some of the specialized practice areas—along with practice-specific standards—have developed in response to settlement pressures. From one or more of these specialized areas of practice, we may find models of possible futures for the broader legal profession.

For example, some sophisticated parties have begun to employ settlement counsel—lawyers who take on a role parallel and complimentary to, but separate from, the role played by the client’s litigators.203 By insulating themselves from some of the duties associated with traditional litigation, settlement counsel are charged with helping clients to overcome some of the most persistent barriers to the efficient resolution of disputes.204 And settlement counsel may dampen some of the challenges of the information asymmetries and mismatching incentives that mark the litigator-client relationship. Settlement counsel at least provide their clients with an internal marketplace of ideas. They are still subject to the potential for principal-agent tensions, of course. Just as a litigator may provide skewed advice, so may settlement counsel. But the client will be in at least a better position to choose from among two different suggested diagnoses and prescriptions.

A second example of the legal landscape changing to alter the underlying incentive and information problems associated with traditional litigation can be found in the rise of collaborative lawyers.205 Unlike settlement counsel, which can be (and often is) unilateral, collaborative law contemplates practitioners on each side of a case. Similar to settlement counsel in that they restrict the scope of their services ex ante, each collaborative lawyer binds herself to represent her clients for purposes of settlement, but precommits not to represent her client in litigation, if settlement talks fail to resolve the dispute.206 Through limited retention agreements or limited scope agreements, these lawyers and their clients create a condition in which they mutually signal not only a willingness, but also a certain level of eagerness, to bargain collaboratively in order to avoid litigation.207 Furthermore, because of the structure of the disqualification provisions, clients and agents share this interest in early resolution.208 Important questions remain about the juxtaposition of these contractual arrangements and lawyers’ overarching ethical duties. But a client who seeks to retain the services of a collaborative lawyer removes at least the prospect that a self-interested lawyer will prescribe protracted litigation, for example.209

I would predict that settlement malpractice claims against either settlement counsel or collaborative lawyers may (either today or at some point in the foreseeable future) be comparatively easier than similar claims against lawyers in general practice.210 Of course, nothing in collaborative lawyers’ limited representation agreements or in settlement counsel’s professional posture could dislodge foundational tort principles such as the duty to exercise ordinary care in delivering those services. But the idea is that each of these practitioners holds herself out to do something differently than other practitioners. Might an unhappy client find in the limited agreements or in the broader representations about the promise of collaborative law a hook upon which to hang a breach of duty claim that might go unrecognized among the broader population of lawyers? Might an examination of collaborative lawyers’ practices reveal that a particular set of behaviors (for example, certain uncollaborative behaviors) might fall demonstrably outside of the relevant community’s standard practices, such that negligence might be established? Might a client (appropriately) find it easier to criticize the settlement-related advice or conduct of an attorney acting as settlement counsel, given that she was specifically hired for the sole purpose of advising and assisting with settlement? Furthermore, because the community of lawyers providing these services is still comparatively small, one might find greater consistency in the kinds of negotiation practices—the kind of consistency one would need in order to demonstrate that a particular practice falls outside of the scope of “ordinary care.” Finally, because these models of practice are specific, rather than general, we might expect to see a change to some of the ethical models governing these practices. The challenge of articulating a standard of care—or a governing ethic—becomes easier in a more constrained and specific practice.

To be clear, I am not suggesting that either collaborative lawyers or settlement counsel are engaged in more malpractice than the general lawyer in a settlement context. Indeed, the collaborative law movement can legitimately claim important successes, and the settlement counsel I have observed have been deeply conscientious and thoughtful practitioners. Instead, my point is that clients of such lawyers might have an easier time establishing that a particular set of attorney actions demonstrably constitute a breach of that kind of lawyer’s duties, even if that same action would be unobjectionable among the broader population of attorneys. The fact of specialization may give rise to a more defined standard of care, thus creating a greater prospect of accountability through a malpractice system.

2. Potential lessons from criminal practice.

Plea negotiations between prosecutors and defense attorneys are in many ways the criminal analogs to civil settlement negotiations. There are, of course, structural, legal, and ethical differences between criminal and civil settlements.211 Particularly when viewed from the perspective of the criminal defendant, however, the similarities between the landscape of practice expectations in plea negotiations and civil settlements overwhelm the differences. Criminal defendants rely on their attorneys for competent advice about their litigation prospects and the implications of any potential plea agreement. And criminal defendants rely on attorneys to engage in most, if not all, of the negotiation dance on their behalf. Each of these aspects of the criminal plea bargaining process has a direct analog in civil settlement, and as I have explained above, each raises the prospect of substandard services by an attorney.

One potentially significant difference, however, is the fact that criminal clients enjoy a constitutional guarantee of effective representation.212 Most of the early cases testing the scope and nature of these rights focused on criminal attorneys’ litigation conduct, but the last decade has seen the Supreme Court also consider the centrality of lawyers’ roles in the plea bargaining process through the lens of the constitutional protection against the ineffective assistance of counsel. For example, in Padilla v Kentucky,213 Jose Padilla, charged with marijuana trafficking,214 received “patently incorrect advice” from his attorney that he accept a plea deal on the grounds that he “did not have to worry about immigration status since he had been in the country so long.”215 In fact, Padilla’s plea made deportation “virtually mandatory.”216 Similarly, in Frye, Galin Frye’s attorney received a plea offer in writing but never communicated the offer to Frye.217 The offer expired, the case went to trial, and Frye was convicted.218 As a final example, in Lafler v Cooper,219 Anthony Cooper was charged with assault with intent to murder.220 He had expressed a willingness to accept a plea agreement initially, but his lawyer (erroneously) informed him that “the prosecution would be unable to establish his intent to murder [the victim] because she had been shot below the waist.”221 Each of these forms of misconduct has direct analogies to the civil settlement misconduct described earlier in this Article. In each of these cases, however, the Supreme Court held that the defendant’s attorney had provided legal services sufficiently substandard to constitute ineffective assistance of counsel.222

Criminal cases raise many of the same questions and challenges for unhappy clients as do their civil counterparts. How does one establish that the lawyer’s error had an actual, adverse impact on the defense?223 How does one establish that the lawyer’s conduct was, in fact, an error?224 How does one overcome the lack of visibility into (or paper trails documenting) the attorney’s advice or negotiation conduct?225 What, if any, impact should a judge’s involvement in the plea bargaining or plea receiving process have on subsequent claims?226 I do not presume that the answers to these questions in the criminal context will be easier than in the civil context—except that the existence of a constitutional protection combined with the near ubiquity of plea agreements will mean that courts are likely to be faced with such claims comparatively more frequently. If nothing else, this volume may, through the usual common law process, stimulate the development of clearer standards about how we judge the conduct of lawyers in a settlement context.227

3. Research prospects.

Settlement negotiations have been the subject of serious scholarly research for only a few generations at most—the blink of an academic eye.228 Some of the indeterminacy associated with negotiation decisions is surely a function of negotiation’s strategic interactivity. But I suggest that we are also on a collective learning curve about negotiation effectiveness. Two different factors will likely drive the next generation of negotiation research: the emergence of big data and a shift away from the search for optimal strategies.

Big data—the collection and analysis of mass volumes of information in ways that were unthinkable a generation ago—has changed countless aspects of modern life,229 and I expect that we will see its effects on our understanding of negotiation as well. What if researchers (or litigants) were able to access broad data about the terms of settlements across jurisdictions and contexts? We have already seen, in the context of certain kinds of personal injury cases, the effects of aggregated data. Big data could offer similar bounding valuation guidance for other types of claims—including those involving settlement.230 Or what if researchers had access to broad data about the process by which settlements were reached? Such information would have the potential to change (or confirm and solidify) our understanding of ordinary (and by extension, extraordinary or outlier) negotiation practices. Having access to the complete record of the flow of offers and other information back and forth in a single negotiation has the potential to yield interesting insights.231 Still, even if the record from a negotiation were complete, data about a single negotiation would not permit the kind of social science to which we might aspire. But what if we had access to that information across multiple years, across thousands of litigated cases, across multiple jurisdictions? What if we had a database of searchable, codable, comparable information about the real-world settlement behaviors (including documents exchanged, information requests, instances of inquiry or advocacy, offers, counteroffers, timelines, and outcomes, for example), and we had that information for thousands, or tens of thousands, or millions232 of parties and their lawyers? Armed with such information, researchers might discover a great deal about how negotiation works in practice. And unhappy parties might discover that their lawyers’ conduct was (or was not) outside of the norm of practice.233

I predict that we will, along with the availability of additional data, see new directions in research. Much of the negotiation research to date has focused largely on the meta question, “What works best?” This research impulse is understandable, and the prescriptive advice stemming from that research is a real improvement over anything that existed even two generations ago. Yet, if theorists are correct, and negotiation almost always presents both opportunities for joint gains and distributive issues to resolve, articulable and absolute prescriptions are almost impossible to construct. “Always make the first offer” is no more defensible as a universal prescription than is “Never make the first offer.” Even if one or the other of those pieces of advice were somehow demonstrably true, presumably both sides would know it to be true, resulting in a tragicomic negotiation stalemate. Particularly with respect to any zero-sum aspect of negotiation, any advice that benefits one side will necessarily come at the expense of the other. And so, both will presumably wind up following the same negotiation advice, rendering its demonstrable effectiveness more questionable.

Even if it is true that it is challenging to articulate universally effective negotiation practices, however, it does not necessarily follow that one could not articulate ineffective negotiation practices. What if researchers—with the benefit of larger datasets—flipped the question modestly? Might there be answers, at least, to the question, “What behaviors tend to work least well in negotiations?” I am unconvinced that we could not identify things that effective negotiators don’t do.234 Perhaps caveats would be necessary. Perhaps even those things we identified would not produce poor results in 100 percent of cases. But even if the results of such an inquiry were merely to identify practices with reliably net expected losses, when compared to some baseline, that would be of great interest. Admittedly, this line of research would likely disappoint busy practitioners looking for advice about what to do next. The market for a popular book entitled, “How Not to Negotiate” would likely be modest. But such a set of research would help to move us beyond the indeterminate vision of negotiation—one in which anything goes. It would instead present a world in which, even if there remains a wide range of different approaches, certain things would fall below the level of “ordinary care,” and from a malpractice perspective, this is all that would be required. Professors Gillian Hadfield and Deborah Rhode have described “validated prescriptive regulations,” based on information exactly like this, as being superior to much of tort law’s current reliance on predictive, performance-based regulation.235 I am fundamentally optimistic that the next generation will know more than us about how disputes are (and are not) resolved effectively.

4. Education as catalyst.

Because tort law compares the actions of a professional with the usual actions of her relevant peer group, the mere identification of best practices is insufficient to cause a meaningful change in the profession’s standards of care. What is required is a shared set of understandings and practices among the members of the profession itself. Legal education provides the greatest possible mechanism for bringing about a widespread change in professional behavior.

Rigorous teaching about negotiation and other forms of settlement is a relatively recent phenomenon. As recently as two generations ago, only a handful of law schools had any courses on the topic.236 Today, many law schools in the United States offer courses on negotiation, mediation, and other nonlitigation dispute resolution mechanisms.237 The ABA has, for several decades, had a Section specifically dedicated to Dispute Resolution. Its membership now exceeds five thousand practitioners, roughly the same size as the ABA’s sections on Intellectual Property, on Criminal Justice, or on International Law, for example.238 There are multiple specialized law journals devoted to the topic,239 and some law schools offer advanced degrees focusing on negotiation or dispute resolution.240 A review of the syllabi of law school courses suggests a remarkable consistency in the negotiation concepts being taught to aspiring lawyers.241 It may have been remarkable at one point that law students were learning to “focus on interests, not positions,” or to focus on understanding and improving a client’s BATNA, for example. A review of several dozen published syllabi and of available casebooks reveals not one example of a law school negotiation course in which these concepts are not taught.242 It is true that negotiation remains an elective—if perpetually oversubscribed—course at most law schools. Still, given the prevalence and consistency of these negotiation lessons, might these (and other) concepts become so familiar, so understood, so adopted as to cause a shift in what constitutes “ordinary care” in the practice of law?

E. What if De Facto Immunity Were Not the Norm?

What policy consequences would follow if the landscape were to shift in a way that caused a change to the de facto immunity lawyers presently enjoy? Concerns tend to fall into three categories: (1) that any reform would drive up the cost of legal services without corresponding benefit, (2) that any reform would dampen the likelihood of settlement, and (3) that reforms would achieve neither appropriate deterrence nor compensation. I acknowledge and address each, briefly, in turn.

Any move toward greater accountability, in any form, creates the real risk that transaction costs will increase. This is true of training requirements, of practice standards, of record keeping, and of virtually anything else that aims to change existing behaviors in any professional context. If more lawyers were exposed to the prospect of accountability in the form of malpractice actions, wouldn’t that added cost simply be passed on to clients? Pointing to medical malpractice premiums, one might reasonably have some concern that lawyers would begin to practice defensive lawyering in a way that drives up costs without corresponding benefit to clients. I am skeptical that the costs of complying with minimal competence standards and documenting that compliance would, in fact, be significant. To the extent patterns of practice change, the costs of engaging in those behaviors goes down. And to the extent there are uncompensated client victims, I am not sure that it is entirely proper to frame appropriate compensation as a “cost.” The question is whether the compensation is appropriate and whether the costs of providing that compensation exceed some other mechanism of deterrence or compensation. Still, I acknowledge that there would surely be some costs associated with making the transition to a system in which the prospect of a settlement malpractice was more than a statistical anomaly.

A second objection I have heard is that any change to lawyers’ roles in the settlement system will cause an undesirable decrease in the rate at which cases settle. Although some question the desirability of consensual resolution as the modal means by which cases are resolved, settlement is unarguably prominent in our current system. I find neither logical nor empirical support, however, for the concern that an increase in lawyers’ exposure to settlement malpractice would cause lawyers to be gun-shy about recommending settlement to their clients. Logically, a lawyer saying, “you should take this offer” is as likely to be flawed as a lawyer saying, “you should reject this offer.” Furthermore, to the extent misconduct is leading to inappropriate settlements (that is, settlements that would not have occurred in a world without misconduct), then as a policy matter, I cannot imagine that we would prefer a system in which those settlements are being deterred.

The third, and in my view, most significant objection to the prospect of expanding lawyers’ exposure to malpractice for their settlement conduct is the risk that tort-based malpractice is the wrong mechanism for achieving the aims of deterrence or compensation. It is certainly true that negligence-based malpractice systems have considerable limitations. Just as the link between medical malpractice cases and quality medical services is debatable,243 perhaps any malpractice regime will sit awkwardly in the context of lawyering and settlement. Furthermore, as Professor David Wilkins has suggested, tort-based liability controls in the lawyering context are challenging because only agency problems that result in large provable damages are likely to be brought into the system. Moreover, litigation against lawyer-defendants is particularly difficult to win. Lawyers are adept at covering their tracks ex ante and fabricating self-interested reconstructions of the facts ex post. In addition, courts tend to be deferential to the exercise of judgment by lawyers. As a result, despite their desire for compensation, clients with small or difficult to prove claims are unlikely to gain access to the malpractice system.244

Perhaps, therefore, we would be better off devoting policy energies toward other approaches to assuring lawyer quality in this context. Perhaps greater education is in order, either as a prelude to professional life, as part of continuing legal education, or as part of the rehabilitation of lawyers who are found to have violated some professional standard.245 Perhaps government officials—either judges or regulators—will gain greater visibility into lawyers’ actions in a way that would permit public accountability.246 Perhaps courts will expand their treatment of such matters through the lens of equity, rather than that of malpractice.247 Still, even if these other avenues are worthwhile, it is inconceivable that a critical part of professional practice (in this case settlement lawyering) should be functionally excluded from the profession’s foundational system of private accountability.


The assumptions at the core of the current legal malpractice system envision lawyers-as-litigators. Even if those assumptions are appropriate for addressing instances of litigation malpractice, their fit with the broader realities of modern practice is awkward. Two facts, juxtaposed, serve to illustrate this dynamic.

First, the vast majority of all civil lawsuits are resolved through negotiated settlements.248 Second, fewer than 1 percent of reported legal malpractice cases and only about 1.5 percent of bar complaints relate to lawyers’ roles in settlement—and even in those instances when clients do complain, lawyers usually prevail.249

I make no empirical claim that lawyers are atypically awful at those aspects of their services that relate to settlement when compared with all other aspects of lawyering. But I see no reason to assume that lawyers are atypically flawless with respect to settlement negotiation and counseling. Instead, I assume that some lawyers sometimes provide substandard advice or other services to their clients in the context of settlement, and I argue that the current malpractice system does a poor job of addressing the modern reality of lawyer-as-settlor.

Malpractice regimes do not, of course, represent the only way to seek to protect clients or to ensure the quality of lawyers’ services. Education, professional norms, ethical standards, heightened governmental scrutiny, and reputational markets also commonly play important roles. But I can envision no overall system of effective professional quality assurance that provides the kind of blanket de facto immunity we see lawyers enjoying with respect to this increasingly critical aspect of their professional service.

Instead, the realities of modern lawyering, in which the lawyer-as-settlor role is more prominent, demand that lawyers be held to a different set of standards than those created solely with litigation in mind. The advice lawyers provide to their clients about the prospect of settlement does not deserve the sweeping judgmental deference of lawyers’ litigation decisions. The case-within-a-case method of judging the impacts of lawyers’ alleged misconduct must adapt, in the settlement context, to the realities of non-zero-sum settlement and probabilistic assessments. Clients’ ability to exercise autonomy—the ability to choose between settlement and continued litigation—should be recognized in practice, not just in the theoretical or aspirational standards of the profession. And we have to be open to the prospect that a lawyer may make a decision in an across-the-table negotiation that is so substandard as to constitute a breach of her professional duty to provide her client with at least ordinary care.

Honest conversation about all of this is critical to the sustained viability of the profession. We should acknowledge the prospect of settlement malpractice, seeking to compensate those who are injured, and seeking to enact structural changes that decrease the risk of such injuries in the future. Negotiated settlements are here to stay. Lawyers will continue to play important roles in those settlements. Clients should be justified in believing not only that their lawyers are improving at this aspect of their practice but also that their lawyers are accountable when they fall short.

Appendix: Research Methodology

A. Reported Malpractice Cases

Online Database.  After verifying that searches in Westlaw and Lexis produce no meaningful differences in results from identical Boolean and other searches, I opted to use Westlaw exclusively.

Date restrictions.  I restricted the search to reported opinions between 2008 and 2017. The data, therefore, include some complaints filed before 2008 and surely excludes some complaints filed in 2017. Ten years represents long enough to produce a sizeable dataset, and I am aware of no intervening changes in law or circumstance that would suggest that I would have seen different results if I had searched 2007 or 2018.

Jurisdictional restrictions.  Although most claims against lawyers are based on state law (tort or contract), complaints against lawyers can arise in either state or federal court. For reasons of research bandwidth, I selected eleven states, rather than conducting a fifty-state survey. The selected states reflect a cross section in terms of (1) region, (2) population, and (3) relevant legal malpractice doctrines, whether stemming from tort or contract-based theories of liability. I excluded states that appear to have relatively idiosyncratic jurisprudence on specific aspects of legal malpractice relevant to settlement malpractice.250

Database filters.  Both Westlaw and Lexis provide prefiltered subdatabases for users wanting to restrict their searches. Westlaw, for example, permits one to restrict searches to include only cases with West Key Number 45 (“Attorney and Client”), and within this, a subcategory of keys related to the “duties and liabilities of attorney to client.” Lexis permits searches to be restricted to their preconstructed “Legal Ethics” database. Neither of these perfectly restricts the dataset to legal malpractice claims, however. These databases include many cases that do not involve an allegation of legal malpractice, as such. For example, these subsets of cases include actions alleging the unauthorized practice of law, as well as disputes about whether state agencies may compel attorneys to provide their social security numbers and whether the attorney or the client must pay for copying costs of materials in a client file following termination.251 These filters also fail to include many legal malpractice actions simply because the issues involved in the particular judicial opinion in question fall outside of their predefined boundaries.252

Search Terms: Legal Malpractice.  Because no existing database fully captures only reported cases about legal malpractice and instead create a risk of under- and overinclusion, and after consulting with research professionals at both Westlaw and Lexis for their guidance,253 I constructed the following Boolean search to use in the full dataset:

(legal lawyer attorney) /5 (malpractice misbehav! offen! violat!)

The results of these searches appear summarized in Table 1 as “Total Reported Legal Malpractice Cases.”

Search Terms: Settlement Malpractice.  Within the broader subset of cases identified as legal malpractice actions, I then sought to isolate those involving alleged settlement malpractice. After conducting a substantive review of several hundred cases from multiple jurisdictions, I concluded that the following Boolean search terms would capture all potential allegations of settlement malpractice:

(Settl! Negot! Offer Compromis! Resol!)

This search produces a smaller dataset of cases, reported in Table 1 as “‘Hits’ Using Search Terms.” Although I am confident that this filter identifies virtually all cases involving settlement malpractice, many of the “Hits” were actually false positives, in the sense that the underlying cases involved allegations of legal malpractice, but the alleged malpractice had nothing to do with settlement malpractice. Some of the false positives were predictable at the outset. For example, settl! and negot! captured a large number of cases related to complaints about the results of plea bargains in the criminal context. Furthermore, there were a number of civil cases in which settlement negotiations were discussed but were not actually the basis of the legal malpractice claim. For example, I read a disturbing number of cases in which attorneys were alleged to have stolen funds from their clients. In many of those cases, the stolen funds were the proceeds of settlements. The alleged impropriety in those cases, however, involved the theft of the money, and there was nothing remarkable about the fact that those funds came from a settlement, rather than a court judgment or from an escrow account, for example.

A sizeable number of false positives also came from the fact that the terms in the search are commonly used in cases not about settlement malpractice—something that could only be sorted out manually on a case-by-case review. settl!, for example, yielded a number of court opinions informing readers that something is well settled under the law.254 negot! described a wide range of different nonsettlement negotiations.255 offer produced cases in which courts ruled against plaintiffs for “offer[ing] no argument challenging the court’s conclusion”256 or for being “unable to offer sufficient expert testimony.”257 compromis! sometimes referred to a party having “compromised his duties,”258 “compromising any confidential information,”259 or even the prospect that sexual relations with a client might “compromise[ ] the client’s legal interests.”260 The resol! search produced a case, for example, in which a court provided an explanation about the timeframes during which post-traumatic stress disorder is expected normally to “resolve” through therapy.261

B. Disciplinary Cases

Date and Jurisdictional restrictions.  As with the civil malpractice search described in the previous Section, I restricted the survey to opinions appearing between 2008 and 2017. I also restricted the research to ten states, aiming for a cross section in terms of (1) region, (2) population, and (3) availability of data. For reasons I describe below, the third of these criteria was particularly important because of the challenges associated with the manner in which many states publish their disciplinary opinions.

Limitation: No Comprehensive, Searchable Databases.  From a research perspective, the most significant challenge in studying disciplinary actions with a focus like this is that no comprehensive dataset exists that would permit the kinds of searches I used with civil malpractice actions.262 Many states publish rolled-up data, for example by providing an overview of the magnitude of disciplinary complaint activity as part of an annual report.263 They may even break those complaints into preexisting categories, but none of them aligned with this Article’s research.264 To the extent searchable databases of disciplinary opinions exist, almost all of them are designed to permit clients or prospective clients to research particular lawyers, rather than to search topics or terms across multiple cases or years.265 I therefore had to construct datasets for each of the states I wished to study. This involved downloading every opinion from every year for each of the selected states. Most states provided these opinions only in a scanned PDF format, requiring individual processing using software to create a searchable document out of each opinion. The result is a unique database of searchable disciplinary opinions requiring several gigabytes of storage for the files.

Search Terms: Settlement Malpractice.  I employed the same Boolean search protocols with the dataset of disciplinary opinions as with civil malpractice actions:

(Settl! Negot! Offer Compromis! Resol!)

As with the earlier search, this produced a number of false positives, for many of the same reasons. As with the civil malpractice search, I found no reliable method for automating the remainder of this search. An individual review of more than one thousand disciplinary opinions identified comparatively fewer false positives in this dataset, probably because, unlike in the Westlaw search, all of these cases necessarily involved an allegation of attorney misconduct of some sort. The only remaining filtering question, therefore, was whether the misconduct was related to settlement.

  • 1As a convention for handling gendered pronouns and in order to maintain consistency and clarity, I make general reference in this Article to clients as he/his/him and attorneys as she/her/hers.
  • 2Each of these is based on allegations made against lawyers in one or more of the reported cases I describe in Part I below.
  • 3Because of methodological differences, no consensus has emerged about the overall civil settlement rate in the United States. Some have claimed settlement rates approaching 98 percent, but I believe that these assertions take the frequency of jury trials (which approaches 2 percent in many contexts) and mistakenly assume that the remainder are resolved through settlement. See, for example, Marc Galanter, A World without Trials, 2006 J Disp Resol 7, 12 (estimating only 1.7 percent of civil cases were resolved by trial in 2006). Instead, some of the remaining cases are dismissed on motion or are abandoned, for example. Still, every credible study of which I am aware has concluded that settlement is at least the modal means by which most forms of civil litigation are resolved. See, for example, Theodore Eisenberg and Charlotte Lanvers, What Is the Settlement Rate and Why Should We Care?, 6 J Empirical Legal Stud 111, 132 (2009) (reporting aggregate settlement rates in the 65–70 percent range); John Barkai and Elizabeth Kent, Let’s Stop Spreading Rumors about Settlement and Litigation: A Comparative Study of Settlement and Litigation in Hawaii Courts, 29 Ohio St J Disp Resol 85, 109 (2014) (reporting roughly similar settlement rates).
  • 4See, for example, Eric W. Orts, Shirking and Sharking: A Legal Theory of the Firm, 16 Yale L & Pol Rev 265, 278 (1998) (describing agency costs in lawyer-client relationships); Robert H. Mnookin, Scott R. Peppet, and Andrew S. Tulumello, Beyond Winning: Negotiating to Create Value in Deals and Disputes 74–76 (Belknap 2000).
  • 5See Michael C. Jensen and William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J Fin Econ 305, 308 (1976).
  • 6Herbert M. Kritzer and Neil Vidmar, When Lawyers Screw Up: Improving Access to Justice for Legal Malpractice Victims 19–36 (Kansas 2018).
  • 7Not all are so circumspect in their critiques of the de facto immunity granted to attorneys under the current laws of certain jurisdictions. See, for example, the full text of Pennsylvania Supreme Court Justice Rolf Larsen’s dissent in Muhammad v Strassburger, McKenna, Messer, Shilobod and Gutnick, in which the majority announced that under Pennsylvania law, clients are barred from recovering in legal malpractice actions in which they settled the underlying litigation—even when the settlement was prompted by the legal malpractice in question:

    The majority has just declared a “LAWYER’S HOLIDAY.” . . . It’s Christmastime for Pennsylvania lawyers. If a doctor is negligent in saving a human life, the doctor pays. If a priest is negligent in saving the spirit of a human, the priest pays. But if a lawyer is negligent in advising his client as to a settlement, the client pays. . . . Thus, “filthy lucre” has a higher priority than human life and/or spirit. The majority calls this “Public Policy.” Maybe . . . Maybe not?? It sure expedites injustice. Should we change the law so that non-lawyers can be judges? I dissent.

    Muhammad v Strassburger, McKenna, Messer, Shilobod and Gutnick, 587 A2d 1346, 1352–53 (Pa 1991) (Larsen dissenting).

  • 8See, for example, Model Rules of Professional Conduct (“Model Rules”) Rule 1.0(e) (ABA 2012) (defining “[i]nformed consent”); Rule 1.2 (“Scope of Representation & Allocation of Authority Between Client & Lawyer”); Rule 1.4 (specifying the required substance of lawyer-client communications).
  • 9Missouri v Frye, 566 US 134, 145 (2012). The Supreme Court has recently taken up the question of lawyer misconduct in the context of plea bargaining, concluding that attorneys’ conduct with respect to plea negotiations is sometimes so deficient as to constitute inadequate representation. See, for example, Lafler v Cooper, 566 US 156, 160, 174 (2012) (concluding that defendant’s attorney conveyed a favorable plea offer to him but imprudently convinced him to reject it based on a flawed articulation of the defendant’s legal risks); Frye, 566 US at 139 (describing that defendant’s attorney failed to convey a plea offer); Lee v United States, 137 S Ct 1958, 1962, 1969 (2017) (explaining that defendant’s attorney advised accepting a plea agreement based on faulty understanding of the legal implications of the plea). Recognizing the centrality of plea negotiations to the modern criminal justice system, the Court has unequivocally recognized that lawyers’ settlement behaviors can fall below the constitutionally required level of care owed to defendants. See note 46.
  • 10Michael Moffitt, Pleadings in the Age of Settlement, 80 Ind L J 727, 771 (2005) (coining the phrase “age of settlement” with respect to the current era of civil litigation).
  • 11The legal profession is not, of course, fully self-governing. See, for example, Fred C. Zacharias, The Myth of Self-Regulation, 93 Minn L Rev 1147, 1153 (2009) (arguing that descriptions of the legal profession as self-regulating are descriptively inaccurate and normatively problematic). Nevertheless, the narrative and values associated with self-governance are well entrenched. See, for example, Model Rules, Preamble ¶ 10 (“The legal profession is largely self-governing.”).
  • 12Lawsuits against attorneys generally sound in tort, contract, or a combination of these two, depending on the jurisdiction. The fundamental components of any claim do not vary meaningfully across these jurisdictions, however. In this Article, I refer to “malpractice,” which suggests an action in tort, but the same analyses would be true in jurisdictions in which courts regard these actions as contract actions. See Merri A. Baldwin, Scott F. Bertschi, and Dylan C. Black, eds, The Law of Lawyers’ Liability ix (ABA 2012) (“While different states impose varying requirements, the basic elements of a legal malpractice claim are essentially the same across the country. A legal malpractice claim can sound in contract or tort or both—except in Alabama, where the only available cause of action is a statutory one.”). States differ in whether they allow tort or contract legal malpractice claims. See, for example, Johnson v Carleton, 765 A2d 571, 573 n 3 (Me 2001) (“tort rather than in contract”); United States National Bank of Oregon v Davies, 548 P2d 966, 968 (Or 1976) (“We have construed such actions to be ones of tort rather than of contract.”); Jackson State Bank v King, 844 P2d 1093, 1095 (Wyo 1993) (“contractual in nature”); Shipman v Kruck, 593 SE2d 319, 322 (Va 2004) (“[A]lthough legal malpractice actions sound in tort, it is the contract that gives rise to the duty.”).
  • 13See Kritzer and Vidmar, When Lawyers Screw Up at 3 (cited in note 6) (providing data on legal services market size); New ABA Data Reveals Rise in Number of US Lawyers, 15 Percent Increase Since 2008 (ABA, May 11, 2018), archived at
    -AY9A (describing data on the number of licensed, active attorneys in the United States).
  • 14Whether eight thousand is a high number, a low number, or just the right number is not my question for this Article. The point, instead, is that one would expect to see some relationship between the frequency with which lawyers undertake an activity and the frequency with which clients complain about that activity. See also Kritzer and Vidmar, When Lawyers Screw Up at 65–72 (cited in note 6) (estimating the frequency of malpractice claims as ranging between 7.5 and 9.7 per 1,000 lawyers, based on data from two national insurance carriers).
  • 15See Table 1.
  • 16Perhaps more honestly, with respect to at least some of these cases, “skimmed.” As I noted in the author’s footnote, I had the benefit of a team of talented research assistants who helped me with many aspects of the empirical work reflected in this Section. Although I reviewed many cases and coded them independently as a double check, I also relied heavily on the judgment of my research assistants with respect to these reviews.
  • 17I observed no difference between the frequency with which settlement malpractice cases appear in published versus unpublished opinions.
  • 18Plaintiffs may, as an empirical matter, have a harder than usual time finding attorneys willing to sue other attorneys. See Kritzer and Vidmar, When Lawyers Screw Up at 144 (cited in note 6) (citing an “absence of a substantial number of plaintiff-side practitioners handling [legal malpractice] cases”).
  • 19I am aware of no data specifically tracking the rate at which legal malpractice cases settle. One study of insurance data indicates that between 56–80 percent of insurance claims involving legal malpractice are resolved without a lawsuit even being filed, suggesting that the percentage of claims that go all the way through trial would be quite low. See id at 28–29. Data from the Standing Committee on Lawyers’ Professional Liability suggest that roughly one quarter of claims against lawyers ripen into lawsuits, and that only about 3 percent of legal malpractice suits then proceed to judgment following trial. Id at 95–96, 125–26. Ronald Mallen reports a trend toward a greater percentage of legal malpractice claims being resolved through settlement, but his data relies principally on studies from 1986 and 1996. See Ronald E. Mallen, 1 Legal Malpractice § 1:18 at 46 (Thomson Reuters 2017) (“[T]here was a significant increase of approximately 8.5 percent in the number of files closed with a settlement payment being made after the commencement of litigation by the claimant.”). Furthermore, common sense and my anecdotal interviews with law firm partners suggest that, at least in corporate counsel settings, firms will commonly adjust bills in the face of unhappy clients—a result that would produce no public paper trail.
  • 20If a client has a conspicuously untenable claim against an attorney, one would expect it to be dismissed early in the life of the litigation and that such a dismissal would be unlikely to merit publication. The dataset of reported cases, therefore, probably includes atypically strong bases for complaints (even if many of them are ultimately unsuccessful).
  • 21In many cases, plaintiffs alleged that the attorneys engaged in multiple acts constituting malpractice. For example, an attorney might be accused of having an undisclosed conflict of interest, of having botched some aspect of the litigation, and of having urged an unwise settlement. In order to be conservative, if any of the alleged conduct involved settlement, the case was counted as involving settlement malpractice.
  • 22Kritzer and Vidmar, When Lawyers Screw Up at 84, 104 (cited in note 6). In their study, Professors Kritzer and Vidmar treated as “successful” any circumstances in which the claimant obtained some payment. This is a different standard than whether the claimant ultimately prevailed at trial. Id at 100.
  • 23Id at 30–31.
  • 24Standing Committee on Lawyers’ Professional Liability, Profile of Legal Malpractice Claims: 2012–2015 15 (ABA 2016). No data were published to permit an assessment of the disposition of just settlement-related conduct.
  • 25See David B. Wilkins, Who Should Regulate Lawyers?, 105 Harv L Rev 799, 805–09 (1992) (suggesting four basic modes of regulating lawyers: disciplinary controls, liability controls, institutional controls, and legislative controls).
  • 26For more on the problematic lack of transparency attached to ethical complaints against lawyers, see generally Leslie C. Levin, The Case for Less Secrecy in Lawyer Discipline, 20 Georgetown J Legal Ethics 1 (2007).
  • 27See id at 17–21 (describing the disciplinary complaint process).
  • 28See, for example, State Bar of Arizona, Annual Report of the Attorney Regulation Advisory Committee to the Arizona Supreme Court (Apr 26, 2018), archived at; The Board of Disciplinary Appeals Appointed by the Supreme Court of Texas, Report 2018, archived at; Oregon State Bar, Disciplinary Board Reporter (2017), archived at; The State Bar of California, Annual Discipline Report for Year Ending December 31, 2018 (Apr 30, 2019), archived at; New York State Bar Association, 2017 Annual Report, archived at
  • 29864 NW2d 794 (Wis 2015).
  • 30Id at 795.
  • 31Id at 798.
  • 32Id.
  • 33Disciplinary Proceedings against Lamb, 864 NW2d at 803. The court ultimately found against Lamb, revoked his license, and ordered him to pay restitution. Id at 805–06.
  • 34Standing Committee on Professional Discipline, 2016 Survey on Lawyer Discipline Systems (S.O.L.D.) *3 (ABA, Jan 2018), archived at (describing 87,487 complaints received by disciplinary agencies and only 3,017 attorneys charged after probable cause determinations—about 3.4 percent of bar complaints).
  • 35For general information on ethics opinions, see Legal Ethics Research Guide: Ethics Opinions (Georgetown Law Library, Oct 16, 2018), archived at
  • 36For a discussion of the ABA’s decision to adopt the Model Rules of Professional Conduct as a replacement for the Model Code of Professional Conduct, see generally Robert J. Kutak, Postscript: The Model Rules of Professional Conduct, 11 Cap U L Rev 585 (1982).
  • 37For example, in Model Rule states, I also searched for and reviewed all opinions about Rules 1.0(e) (“Informed consent”), 1.2(a) (allocation of authority), 1.4(a)(1), and 1.4(b) (communication with client).
  • 38This possibility supports the description of lawyering as a “credence good.” See, for example, Gillian K. Hadfield and Deborah L. Rhode, How to Regulate Legal Services to Promote Access, Innovation, and the Quality of Lawyering, 67 Hastings L J 1191, 1196 (2016) (describing lawyering as a “credence good” because consumers of legal services struggle to “judge the quality of the services provided”); Nathaniel G. Hilger, Why Don’t People Trust Experts?, 59 J L & Econ 293, 306–08 (2016) (explaining the dynamics of markets for credence goods, including legal services). See also John C.P. Goldberg, What Are We Reforming? Tort Theory’s Place in Debates over Malpractice Reform, 59 Vand L Rev 1075, 1078 (2006) (discussing how plaintiffs may refrain from bringing medical malpractice suits due to information asymmetries); Kritzer and Vidmar, When Lawyers Screw Up at 12 (cited in note 6) (describing the “[r]ecognition and [a]ttribution [p]roblem” for plaintiffs bringing professional malpractice suits).
  • 39See, for example, Jennifer K. Robbennolt and Jean R. Sternlight, Psychology for Lawyers 21–24 (ABA 2012) (describing cognitive dissonance and its frequent coupling with confirmation biases).
  • 40See Kritzer and Vidmar, When Lawyers Screw Up at 144 (cited in note 6).
  • 41Of all of the possible explanations I offer in this paragraph, this final one is the only one I am comfortable rejecting outright. Nothing in my experience as a law professor, as a mediator, or as a consultant in negotiations supports the picture of lawyers as infallible negotiators.
  • 42See Part I.A.
  • 43See Part I.B.
  • 44See Part I.C.
  • 45The line between lawyer-as-litigator and lawyer-as-settlor is not a bright one, of course. Settlement counseling and settlement negotiations take place “in the shadow of the law,” with expected litigation outcomes being among the principal drivers of the relevant legal endowments and constraints. Robert H. Mnookin and Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce, 88 Yale L J 950, 968 (1979).
  • 46See Ronald E. Mallen, 2 Legal Malpractice § 20:2 at 1321 (Thomson Reuters 2017) (“[A]n attorney should exercise the skill and knowledge ordinarily possessed by attorneys under similar circumstances.”) (emphasis in original). In the context of plea bargaining, recent Supreme Court opinions also make clear that there is a constitutional aspect to lawyers’ duties to their clients—what Professor Rishi Batra called “a negotiation competency bar for criminal defense attorneys.” Rishi Batra, Lafler and Frye: A New Constitutional Standard for Negotiation, 14 Cardozo J Conflict Resol 309, 310 (2013). I would imagine that this standard is lower than the bar of “ordinary care.” That is, I presume that attorneys as a group do not “ordinarily” fall below the constitutionally mandated minimum effectiveness. See Cynthia Alkon, Plea Bargain Negotiations: Defining Competence beyond Lafler and Frye, 53 Am Crim L Rev 377, 399–406 (2016) (arguing for competence stan-dards in plea bargaining beyond the context of client counseling).
  • 47Because “ordinary care” is context specific, no exhaustive list of per se breaches of such care exists. Examples from the litigation context include such things as “procedural missteps that foreclosed opportunities to take advantage of legal remedies, such as failure to serve process, negligent conduct of a trial, failure to attend the trial, or failure to call crucial witnesses.” Zalta v Billips, 81 Cal App 3d 183, 188 (1978). Whether particular conduct constitutes a breach of “ordinary care” is generally a matter in which the factfinder must rely on expert testimony. See, for example, First National Bank of LaGrange v Lowrey, 872 NE2d 447, 464 (Ill App 2007).
  • 48See Mallen, 2 Legal Malpractice § 19:1 at 1232 (cited in note 46). See also id § 33:85 at 924 (discussing how lawyers’ judgment in settlement decisions is generally given deferential treatment); State v Madison, 770 P2d 662, 667 (Wash App 1989) (“The decision of when or whether to object is a classic example of trial tactics. Only in egregious circumstances . . . will the failure to object constitute incompetence of counsel.”).
  • 49See Mallen, 2 Legal Malpractice § 20:2 at 1320 (cited in note 46).
  • 50The confidentiality attached to most legal negotiations creates a challenge even for expert testimony to reliably establish practice norms with respect to specific negotiation conduct.
  • 51566 US 134 (2012).
  • 52Id at 145.
  • 53I have been teaching law students and consulting with practicing attorneys, executives, and diplomats in executive training programs for more than two decades. I commonly hear them say things like, “I could never negotiate the way my [former boss/best friend/uncle/mentor] negotiates, even though she/he seems to have success with it.”
  • 54For example, Professors Deborah Kolb and Linda Putnam describe a “double bind” for women in negotiation because some of the behaviors traditionally associated with negotiation success “when enacted by a woman are likely to be seen differently than they are when men employ them.” Deborah M. Kolb and Linda L. Putnam, Negotiation through a Gender Lens, in Michael L. Moffitt and Robert C. Bordone, eds, The Handbook of Dispute Resolution 137–38 (Jossey-Bass 2005). Similarly, one’s contextual “social positioning” may have a serious effect on how one’s negotiation actions are received. See Carol Watson, Gender versus Power as a Predictor of Negotiation Behavior and Outcomes, 10 Negot J 117, 120 (1994) (explaining that “situational power” may influence how men and women behave in negotiating contexts). See also generally Hannah Riley and Kathleen L. McGinn, When Does Gender Matter in Negotiation? (HKS Faculty Research Working Paper Series, Sept 2002), archived at (analyzing the material effects of gender in negotiations).
  • 55One might further argue that, as a policy matter, failing to defer to this kind of individual choice might risk calcifying a particular set of existing practices, precluding the eventual development of even more effective practices across the field.
  • 56Both as to parallels to the ethics of negotiation and as to the risk-reward calculations and associated information asymmetries, many have suggested a connection between poker and negotiation. See, for example, How to Play Your Hand: Lessons for Negotiators from Poker, 2 UNLV Gaming L J 231, 234, 248 (2011); Russell Korobkin, Michael Moffitt, and Nancy Welsh, The Law of Bargaining, 87 Marq L Rev 839, 839 (2004). See also generally Steven Lubet, Lawyers’ Poker: 52 Lessons that Lawyers Can Learn from Card Players (Oxford 2006).
  • 57For example, “judgmental immunity” is often applied in roughly the same manner in the execution of settlement agreements as with tactical litigation decisions. See, for example, Meyer v Wagner, 709 NE2d 784, 786, 791 (Mass 1999) (no malpractice liability for pursuing reasonable strategies in executing a settlement agreement that ultimately failed).
  • 582002 WL 31772008 (Cal App).
  • 59Id at *1.
  • 60Id at *2–3.
  • 61Id at *3–4.
  • 62Ram, 2002 WL 31772008 at *13, citing California Rules of Professional Conduct, Rule 3-700(C)(1)(e). Taken to its extreme, this logic must have limits. We would surely perceive a role for the client in choosing between Litigation Strategy A (giving her an 80 percent chance of winning $1 million—an $800,000 expected litigation value) and Litigation Strategy B (giving her a 10 percent chance of winning $8 million—also an $800,000 expected litigation value). Still, courts have routinely deferred to lawyers’ roles in negotiation when their judgment clashed with that of their clients.
  • 63See generally Howard Raiffa, The Art and Science of Negotiation (Belknap 1982).
  • 64For a survey of some of the conflicting literature in this regard, see Robbennolt and Sternlight, Psychology for Lawyers 273–74 (cited in note 39).
  • 65I thank Professor Scott Peppet for suggesting this analogy in this context. For a detailed look at the analogy between jazz and settlement mediation, see generally John W. Cooley, Mediation, Improvisation, and All That Jazz, 2007 J Disp Resol 325.
  • 66See Michael Wheeler, The Art of Negotiation: How to Improvise Agreement in a Chaotic World 97–117 (Simon & Schuster 2013). See also generally Jeffrey Krivis, Improvisational Negotiation: A Mediator’s Stories about Love, Money, and Anger—and the Strategies That Resolved Them (Jossey-Bass 2006).
  • 67Mallen, 1 Legal Malpractice § 8:20 at 1037–41 (cited in note 19) (discussing the “substantial factor” test for establishing causation in legal malpractice claims).
  • 68Robert E. Mallen, 4 Legal Malpractice § 37:1 at 1446 (Thomson Reuters 2017). See also Restatement (Third) of the Law Governing Lawyers § 53 (2000) (explaining causation and damages for lawyer civil liability).
  • 69“Would have” and “should have” are typically treated as the same question in this context. See Mallen, 4 Legal Malpractice § 37:87 at 1700–01 (cited in note 68).
  • 70See Randall Kiser, Beyond Right and Wrong: The Power of Effective Decision Making for Attorneys and Clients 223 (Springer 2010) (“Although attorneys uniformly acknowledge the principle that only the client can consent to a settlement, this principle is strained in actual practice.”).
  • 71This presumption is rebuttable in a malpractice action, in a third-party action, or in an action to set aside the settlement. However, in any of those cases, the party seeking to demonstrate that the attorney lacked the authority will have the burden of producing evidence to support that assertion. See generally, for example, In re Artha Management, Inc, 91 F3d 326 (2d Cir 1996). See also Surety Insurance Co of California v Williams, 729 F2d 581, 582–83 (8th Cir 1984) (“Although an attorney is presumed to possess authority to act on behalf of the client, a judgement entered upon an agreement by the attorney may be set aside on affirmative proof that the attorney had no right to consent to its entry.”) (quotation marks omitted), quoting Bradford Exchange v Trein’s Exchange, 600 F2d 99, 102 (7th Cir 1979); St. Amand v Marriott Hotel, Inc, 430 F Supp 488, 490 (ED La 1977):

    The law is settled that an attorney of record may not compromise, settle or consent to a final disposition of his client’s case without express authority . . . . However, this general principle must be considered in connection with the rule that an attorney of record is presumed to have authority to compromise and settle litigation of his client, and a judgment entered upon an agreement by the attorney of record will be set aside only upon affirmative proof of the party seeking to vacate the judgment that the attorney had no right to consent to its entry.

    (quotation marks omitted). In some cases, of course, those seeking to challenge the settlement are able to meet this burden and demonstrate that an attorney lacked the authority to settle. See, for example, Blutcher v EHS Trinity Hospital, 746 NE2d 863, 872 (Ill App 2001).

  • 72587 A2d 1346 (Pa 1991).
  • 73Id at 1348.
  • 74Regrettably, the Pennsylvania Supreme Court did not acknowledge the distinction between the case’s litigation value and its settlement value.
  • 75Wolski v Wandel, 746 NW2d 143, 149 (Neb 2008). See also Ziegelheim v Apollo, 607 A2d 1298, 1304 (NJ 1992) (rejecting the categorical application of the rule espoused in Muhammad); Grayson v Wofsey, Rosen, Kweskin and Kuriansky, 646 A2d 195, 199–200 (Conn 1994); Bill Branch Chevrolet, Inc v Burnett, 555 S2d 455, 456 (Fla App 1990) (“We cannot say as a matter of law that the settlement of this case negates any alleged legal malpractice as a proximate cause of loss.”); Pike v Mullikin, 965 A2d 987, 991 (NH 2009) (spouse who consented to resolution of contested divorce not precluded from bringing subsequent malpractice action against attorney for her role in drafting the contested antenuptial agreement). Even in Pennsylvania, the Third Circuit declined to extend the Muhammad bar in a failure-to-prosecute case. See Wassal v DeCaro, 91 F3d 443, 446 (3d Cir 1996).
  • 76See, for example, Elmo v Callahan, 2012 WL 3669010, *8 n 10 (D NH).
  • 7772 Cal App 4th 1514 (1999).
  • 78Id at 1516.
  • 79Id at 1519 (emphasis in original).
  • 80See, for example, Whiteaker v State, 382 NW2d 112, 115–17 (Iowa 1986) (requiring plaintiff to provide evidence of the specific terms of a settlement that was neither made nor offered in order to overcome proximate causation burden in a malpractice case).
  • 81518 SW3d 394 (Tex 2017).
  • 82Id at 398–99.
  • 83Id at 410–11.
  • 84Id at 411.
  • 85Part of the challenge was that Rogers would have needed to prove not only that he would have accepted the offer, but also that he would have had funds sufficient to make the payment in question. Rogers, 518 SW3d at 411.
  • 86Id at 411.
  • 87Thomas v Bethea, 718 A2d 1187, 1197 (Md 1998). These same challenges arise in the context of alleged attorney misconduct in the context of transactional dealmaking. See, for example, Viner v Sweet, 70 P3d 1046, 1054 (Cal 2003) (“[J]ust as in litigation malpractice actions, a plaintiff in a transactional malpractice action must show that but for the alleged malpractice, it is more likely than not that the plaintiff would have obtained a more favorable result.”).
  • 88In some circumstances, the intervening decisions may be those of another attorney—for example, because the client fired his first one. See, for example, Meiners v Fortson & White, 436 SE2d 780, 781 (Ga App 1993) (explaining that a first lawyer made errors, but was shielded from liability on proximate causation grounds because a second attorney made subsequent errors); Royal Insurance Co of America v Miles & Stockbridge, PC, 138 F Supp 2d 695, 698 (D Md 2001) (compounding errors by series of attorneys, creating what the court described as a “megaplex of errors already committed by virtually everyone who had come close to this nettlesome mess”).
  • 89For a survey of the contexts in which courts commonly review the substantive terms of settlement agreements, see Sanford L. Weisburst, Judicial Review of Settlements and Consent Decrees: An Economic Analysis, 28 J Legal Stud 55, 77, 81, 83, 92 (1999).
  • 90See, for example, Meyer v Wagner, 709 NE2d 784, 791 (Mass 1999); Ex parte Free, 910 S2d 753, 756 (Ala 2005) (concluding that plaintiff’s failure to contest a settlement agreement was immaterial when the complaint alleged that information relating to the settlement was misrepresented).
  • 91872 NE2d 447 (Ill App 2007).
  • 92Id at 455.
  • 93Id at 455–56.
  • 94This creates a curious litigation posture for defendants like the hospital in this case. At trial, they argued that their own settlement offer of $1 million was not nearly enough to cover the plaintiff’s son’s expected medical expenses. Id at 467.
  • 95The client might also have a claim for any unrecovered transaction costs associated with the ensuing litigation because the litigation was unnecessarily caused by the attorney’s misconduct.
  • 96159 AD3d 683 (NY App 2018).
  • 97Id at 683.
  • 98Id at 684.
  • 99See, for example, Whiteaker, 382 NW2d at 115–17.
  • 100211 Cal App 4th 154 (2012).
  • 101Id at 166. In Filbin, the plaintiffs sought to demonstrate that they would have obtained a more favorable settlement but for their attorney’s mistaken advice in an eminent domain case. The court rejected as too speculative the Filbins’ assertion that the other party would have settled for a more favorable amount but for the attorney’s error. Id at 171. The negotiation literature broadly supports this hesitation to speculate about such matters. Outside of exceptionally narrow conditions, one cannot know whether a Zone of Possible Agreement (ZOPA) exists. Even if one somehow knew such a zone’s precise contours, one would still not know whether agreement would result, much less agreement on any particular terms. See generally Katie Shonk, How to Find the ZOPA in Business Negotiations (Program on Negotiation at Harvard Law School, Oct 18, 2018), archived at
  • 102The Chemical Safety Board’s report says 15 killed and 180 injured. US Chemical Safety and Hazard Investigation Board, BP America Refinery Explosion (Chemical Safety Board, Mar 20, 2007), archived at
  • 103Elizondo v Krist, 415 SW3d 259, 260, 263 (Tex 2013).
  • 104Id at 260, 261.
  • 105Id at 263. Elizondo’s claim was dismissed on summary judgment, however, because the Texas Supreme Court judged the expert’s assessment of the relationship between this aggregated settlement data and the plaintiff’s settlement to be “too conclusory.” Id at 265. But see Fishman v Brooks, 487 NE2d 1377, 1380 (Mass 1986) (permitting expert testimony as “evidence of the fair settlement value of the underlying claim”).
  • 106See, for example, Biz Carson, One of Peter Thiel’s Fellows Created a New Startup That Will Fund Your Lawsuit (Business Insider, Aug 24, 2016), archived at; Business Solutions: Manage Legal Cost and Risk (Burford 2019), archived at
  • 107See Jonathan T. Molot, Litigation Finance: A Market Solution to a Procedural Problem, 99 Georgetown L J 65, 101–04 (2010) (discussing litigation financing as a means toward a more robust litigation market); Susan Lorde Martin, Litigation Financing: Another Subprime Industry That Has a Place in the United States Market, 53 Vill L Rev 83, 102 (2008). See also generally Blakeley B. McShane, et al, Predicting Securities Fraud Settlements and Amounts: A Hierarchical Bayesian Model of Federal Securities Class Action Lawsuits, 9 J Empirical Legal Stud 482 (2012) (developing models to predict incidence and amounts of securities fraud settlements).
  • 108See, for example, Charlotte Alexander, Using Analytics to Detect Legal Risk (Brink, May 8, 2018), archived at (finding an artificial intelligence tool 20 percent more accurate than judgments of 100 lawyers from top London firms).
  • 109See, for example, Capitalize on Intelligence from Prior Case Outcomes: LexisNexis Verdict & Settlement Analyzer: Case Assessment and Planning (LexisNexis 2010), archived at; Westlaw Case Evaluator (Thomson Reuters), archived at
  • 110One interesting analytical approach was suggested in Glenna v Sullivan, 245 NW2d 869 (Minn 1976). The Minnesota Supreme Court affirmed the trial court’s decision to direct a verdict against a plaintiff who had alleged that her attorney had negligently prepared for her case and had, therefore, recommended an inadequate settlement. Id at 869, 873. Justice John J. Todd, concurring with the decision to dismiss the appeal, examined the substance of the settlement and assessed it as though it were a jury verdict being reviewed for additur purposes, writing, “The settlement amount, in my judgment, represents the low range of jury verdict which would be sustained without additur if the case had been properly prepared and tried.” Id at 873 (Todd concurring in the judgment).
  • 111If lawyers’ settlement duties were so nebulous that it was impossible to distinguish competence from incompetence, then clearly deference—or even blanket immunity—might be required. As I explain below, however, there are at least some settlement-related duties that are so universally recognized as to permit examination and a reasonable expectation that those duties will be satisfied.
  • 112Roger Fisher and William Ury, Getting to Yes 42 (Houghton Mifflin 1981). See also Mnookin, Peppet, and Tulumello, Beyond Winning at 37 (cited in note 4) (“[U]nderlying interests are the stuff of which value-creating trades are made.”); Robbennolt and Sternlight, Psychology for Lawyers at 256–60 (cited in note 39) (giving an overview of some of the real-world challenges of assessing client interests). I treat interests here as a slightly broader category than the Model Rules’ dictates about the “objectives of representation,” because I have heard some suggest that “of representation” creates a narrowing of considerations. A client has interests that may or may not be related to a specific representation only by virtue of whether litigation takes a particular turn, or by whether a creative clause is inserted into a settlement.
  • 113Model Rules, Rule 1.2(a).
  • 114See Fisher and Ury, Getting to Yes at 59 (cited in note 112).
  • 115Even in a litigation context, a lawyer’s role properly extends beyond this simple duty. A lawyer whose client faces a high-profile lawsuit may coordinate with a crisis communication team, for example, to help the client to understand the risks and opportunities associated with actions that may or may not have any effect on either settlement or litigation. A lawyer whose client is in need of social services or other resources may find ways to help him access those services in ways that have no effect on the immediate litigation or its settlement. See Model Rules, Rule 1.0(e) (discussing communication of material risks).
  • 116See, for example, Model Rules, Rules 1.2(a), 1.4(a)(2), 1.4(b) (describing the scope of representation and a lawyer’s responsibility to communicate with the client).
  • 117Not all cases of legal malpractice will be violations of legal ethics and vice versa. See Mallen, 1 Legal Malpractice § 1:22 at 53 (cited in note 19); Mallen, 2 Legal Malpractice § 20:9 at 1358–59 (cited in note 46).
  • 118And when clients have multiple interests, a lawyer needs to know the relative strengths of those interests. Otherwise, she will be left to make judgments on behalf of the client—judgments that more properly lie with the client. See, for example, Arden v Forsberg & Umlauf, 373 P3d 320, 334–35 (Wash App 2016) (explaining that the client had personal as well as pecuniary interests in a case, leaving the lawyer to exercise judgment about those interests’ relative weight or importance in settlement negotiations).
  • 119See Model Rules, Rule 1.4.
  • 120751 NE2d 936 (NY 2001).
  • 121The original agreement had noted the settlement amount as $2,800,000 rather than $2,080,000. Id at 937.
  • 122Id.
  • 123Id at 938.
  • 124But see Berman v Rubin, 227 SE2d 802, 804 (Ga App 1976) (concluding that a plaintiff who read and accepted terms of settlement has no case for legal malpractice against an attorney who provided inadequate explanation of those settlement terms).
  • 125Even as to this, lawyers must sometimes take great care in describing the implications of deal terms to their clients. See, for example, Gulliver Schools, Inc v Snay, 137 S3d 1045, 1046 (Fla App 2014). As part of the settlement of an age discrimination suit, Patrick Snay agreed to a confidentiality clause providing that he would not communicate even the existence of a settlement, much less its terms, to anyone other than his attorneys or other professional advisors. Snay informed his teenage daughter that the case had been settled and that he was happy. Id. His daughter then posted a message to social media reading, “Mama and Papa Snay won the case against Gulliver. Gulliver is now officially paying for my vacation to Europe this summer. SUCK IT” (teenage caps in original). The court treated this as a breach and refused to compel Gulliver to pay Snay the settlement. Id at 1046. Surely a lawyer might have been able to help Snay to consider the implications of the wording of the confidentiality clause.
  • 126425 F Supp 2d 276 (D Conn 2006).
  • 127Id at 277.
  • 128Id at 277–78.
  • 129Id at 278.
  • 130See Ordon, 425 F Supp 2d at 278. Ordon ultimately lost his legal malpractice claim against the attorney who provided this advice because he was unable to prove, through expert testimony, that he would have prevailed in the action before the Connecticut board. Id at 280–82. The court reasoned that Ordon would have suffered the same extrajurisdictional fate if he had lost before the medical board, so unless he could prove that he would have won in that action, Ordon could not establish proximate causation. Id.
  • 131624 A2d 117 (Pa Super 1993).
  • 132Id at 119.
  • 133Id.
  • 134Id at 119–21.
  • 135In the simplest of contexts, the lawyer is helping her client to make a binary choice: to settle on these terms or to resolve the case through litigation. As a practical matter, clients often perceive the choice to be more complex: to settle on these terms, or to continue with litigation, while still holding out the hope that a more attractive settlement offer may emerge later in the course of litigation.
  • 136See Fisher and Ury, Getting to Yes at 99–108 (cited in note 112).
  • 137Sauer v Flanagan and Maniotis, 748 S2d 1079, 1080 (Fla App 2000) (the lawyer also said the client would be “crazy” to accept a million-dollar settlement offer). In her subsequent legal malpractice action, the plaintiff survived summary judgment, and the case was settled in mediation, with the terms of the settlement filed under seal. See Sauer v Flanagan and Maniotis, Docket No 237 (Fla App filed Sept 5, 2000). But see Miranda v Said, 836 NW2d 8, 11, 13 (Iowa 2013) (attorney told clients his proposed course of action for entry into the United States presented “no risks and had a ninety-nine percent chance of success” despite obstacles that experts later opined made the plan legally untenable).
  • 138See Grayson v Wofsey, Rosen, Kweskin and Kuriansky, 646 A2d 195, 203 (Conn 1994).
  • 139See Thomas v Bethea, 718 A2d 1187, 1195–96 (Md 1998).
  • 140See Charnay v Cobert, 145 Cal App 4th 170, 176 (2006).
  • 141The analytic approach underlying LRA coalesced in its current form in 1964 with the work of Professors Howard Raiffa and Ronald Howard, and the term “decision analysis” was coined. See Justin Fox, How to Tell if You’ve Made a Good Decision (Harvard Business Review, Nov 21, 2014), archived at (describing Raiffa and Howard as cofounders of the “[d]ecision analysis” field). See also generally Marjorie Corman Aaron, Risks & Rigor: A Lawyer’s Guide to Decision Trees for Assessing Cases and Advising Clients (DRI 2018); Raiffa, The Art and Science of Negotiation (cited in note 63); David P. Hoffer, Decision Analysis as a Mediator’s Tool, 1 Harv Negotiation L Rev 113 (1996); Marjorie Corman Aaron and David P. Hoffer, Decision Analysis as a Method of Evaluating the Trial Alternative, in Dwight Golann, ed, Mediating Legal Disputes 307–34 (Little, Brown 1996); John S. Hammond, Ralph L. Keeney, and Howard Raiffa, Smart Choices: A Practical Guide to Making Better Decisions (Harvard Business School 2002).
  • 142See generally, for example, Hoffer, 1 Harv Negotiation L Rev (cited in note 141); Jeffrey M. Senger, Decision Analysis in Negotiation, 87 Marq L Rev 723 (2004).
  • 143Marjorie Corman Aaron, Finding Settlement with Numbers, Maps, and Trees, in Moffitt and Bordone, The Handbook of Dispute Resolution at 204 (cited in note 54).
  • 144Id at 205.
  • 145For a useful overview of these difficulties, see Mark K. Osbeck, Lawyer as Soothsayer: Exploring the Important Role of Outcome Prediction in the Practice of Law *66 (Michigan Law & Econ Working Paper, Mar 12, 2018), archived at

    Outcome prediction has always been a vital part of practicing law. Clients of all types rely on their attorneys to provide accurate assessments of the potential legal consequences the clients face when making important decisions. And yet, notwithstanding its enormous importance to the practice of law, outcome prediction in the law remains a very imprecise endeavor.

    See also generally Daniel Martin Katz, Quantitative Legal Prediction–Or–How I Learned to Stop Worrying and Start Preparing for the Data-Driven Future of the Legal Services Industry, 62 Emory L J 909 (2013).

  • 146See, for example, Jane Goodman-Delahunty, et al, Insightful or Wishful: Lawyers’ Ability to Predict Case Outcomes, 16 Psychology Pub Pol & Law 133, 139–48 (2010).
  • 147There is some evidence, for example, that lawyers who consult with another colleague produce more reliable predictions than those who rely on their own assessments. Edie Green and Brian Bornstein, Cloudy Forecasts, 47-Apr Trial 28, 31–32 (2011).
  • 148See, for example, Theodore W. Ruger, et al, The Supreme Court Forecasting Project: Legal and Political Science Approaches to Predicting Supreme Court Decisionmaking, 104 Colum L Rev 1150, 1171 (2004) (showing that a statistical model outperforms a panel of legal experts in predicting Supreme Court cases).
  • 149Perhaps state offer of judgment statutes might provide some guidance about the magnitude of an appropriate buffer. The basic structure of offer of judgment statutes is that a party who rejects a qualifying offer and fails to achieve a better outcome at trial will be liable to the offering party for certain expenses incurred after the offer was made. Some states provide a cushion in these assessments. See, for example, Alaska Stat Ann § 09.30.065 (giving a 5 percent cushion); 2 Mich Ct Rules Prac, Rule 2.403(O)(3) (giving a 10 percent cushion); Fla Stat § 768.79 (2018) (giving a 25 percent cushion). I fear that our lawyer-driven system would likely choose to hold attorneys to a lesser level of precision in their predictions, even though we punish clients for “unreasonably” rejecting offers outside of that interval.
  • 150See, for example, Jones v Lattimer, 29 F Supp 3d 5, 13 (DDC 2014) (finding triable fact in which the client said that her attorney “never advised [her] of the risk associated with turning down” settlement offers of $3 million and $1 million); Clark v Bristol-Myers Squibb & Co, 306 AD2d 82, 84 (NY App 2003) (finding triable issue of fact whether the plaintiff in a silicone breast implant case had authorized her attorney to agree to settlement on a particular set of terms); Bastys v Rothschild, 154 Fed Appx 260, 263 (2d Cir 2005) (finding that plaintiff alleged that attorney’s advice to settle a divorce action was negligent, but dismissing the claim on summary judgment grounds because the plaintiff could not demonstrate that the attorney-defendant had given that advice).
  • 151The demise (or at least decline) of lengthy, formal client memoranda is not recent. See, for example, Kristen Konrad Robbins-Tiscone, From Snail Mail to E-mail: The Traditional Legal Memorandum in the Twenty-First Century, 58 J Legal Educ 32, 34–35 (2008) (describing survey results showing a reduced use of formal legal memoranda in the practice of law). A paper trail need not be so onerous or costly as these memos, however.
  • 152Compare Susan R. Martyn, In Defense of Client-Lawyer Confidentiality . . . and Its Exceptions . . ., 81 Neb L Rev 1320, 1321 (2002), with Louis Kaplow and Stephen Shavell, Legal Advice about Information to Present in Litigation: Its Effects and Social Desirability, 102 Harv L Rev 565, 608–13 (1989).
  • 153See generally Atul Gawande, The Checklist Manifesto: How to Get Things Right (Metropolitan Books 2009).
  • 154Id at 80–81.
  • 155Id at 154.
  • 156Id at Appx 4 (“A checklist is NOT a teaching tool or an algorithm.”).
  • 157An additional benefit to written records is that they can, in some circumstances, facilitate clearer communication when the client’s native language does not match that of the lawyer. In Ram v Cooper, for example, the client wrote his attorney, saying, “Since, I am not well versed in English Language, I prefer all communication to be in writing to prevent any misconception in any manner or form. What is there that you can only say in private and not in writing?” Ram, 2002 WL 31772008 at *9.
  • 158One interesting possibility would be that if written advice were the professional norm, we might see courts shift the presumptions or burdens in a legal malpractice action in which there was no written record of the advice. In other words, perhaps the thumb would be on the scale in favor of the complaining client if his attorney cannot produce a record of the advice.
  • 159One might look to contract provisions such as license agreements or arbitration clauses for a cautionary example in this regard. See Jeff Sovern, et al, “Whimsy Little Contracts” with Unexpected Consequences: An Empirical Analysis of Consumer Understanding of Arbitration Agreements, 75 Md L Rev 1, 43–62 (2015) (using empirical methods to find prevalent and profound misunderstandings of arbitration clauses); Yannis Bakos, Florencia Marotta-Wurgler, and David R. Trossen, Does Anyone Read the Fine Print? Consumer Attention to Standard-Form Contracts, 43 J Legal Stud 1, 32 (2014) (finding that less than 0.2 percent of consumers read online end user license agreements with any care). But neither of those purports to be an encapsulation of advice from one with a fiduciary duty to the recipient. The better, although still perhaps cautionary, model might be the medical profession, in which we still sometimes see consent forms bordering on the meaningless. See, for example, Melissa M. Bottrell, et al, Hospital Informed Consent for Procedure Forms: Facilitating Quality Patient-Physician Interaction, 135 Archives Surgery 26, 29 (2000) (finding only 26.4 percent of hospital forms contained all basic elements of informed consent).
  • 160There are contexts in which an attorney’s oral advice might mismatch her written advice. Recall, for example, Sauer, in which the attorney allegedly assured the client that he had a 99 percent chance of success, but refused to document that assessment in writing. Sauer, 748 S2d at 1080.
  • 161This ignores the reality that a litigator’s substandard decisions are rarely, in fact, dispositive by themselves. Instead, most substandard decisions within a litigation context merely decrease the client’s likelihood of prevailing. But the current system, relying on a case-within-a-case approach, recognizes only a subset of lawyer misconduct as injurious—and then often overcompensates clients in those instances.
  • 162In the real world, neither of these conditions is likely. See, for example, Randall Kiser, Beyond Right and Wrong at 20–24 (cited in note 70); Lucian Arye Bebchuk, Litigation and Settlement under Imperfect Information, 15 RAND J Econ 404, 406–09 (1984).
  • 163Assume, for example, that the lawyer failed to give notice of an expert witness, and as a result, will have a much harder (but not impossible) time prevailing on one of the necessary elements of the plaintiff’s claim.
  • 164Recall that the case-within-a-case methodology relies on a preponderance of the evidence standard. Mallen, 4 Legal Malpractice § 37:97 at 1730 (cited in note 68) (“evidence that, more likely than not, the attorney’s conduct caused injury”). It would not differentiate between an 80 percent case, a 55 percent case, and a 99.44 percent case.
  • 165For more on the distinction between “moral wrongs” and “legal wrongs,” see John C.P. Goldberg and Benjamin C. Zipursky, Torts as Wrongs, 88 Tex L Rev 917, 930–32 (2010). For more discussion of the difference between causation and the value of a claim, see Joseph H. King Jr, Causation, Valuation, and Chance in Personal Injury Torts Involving Preexisting Conditions and Future Consequences, 90 Yale L J 1353, 1354 (1981) (“Courts have had difficulty perceiving that a chance of avoiding some adverse result or of achieving some favorable result is a compensable interest in its own right.”).
  • 166If a doctor fails to diagnose a patient’s cancer, and a properly diagnosed patient would have had a 30 percent chance of recovery, the patient has a cause of action against his doctor only in some jurisdictions. See, for example, Matsuyama v Birnbaum, 890 NE2d 819, 842 (Mass 2008) (allowing recovery when the defendant doctor’s negligence was the but-for cause of patient’s “loss of chance”); Roberts v Ohio Permanente Medical Group, Inc, 668 NE2d 480, 482 (Ohio 1996) (recognizing claims for “loss of chance” in wrongful death actions). The loss of chance doctrine has not, however, as a general rule, extended to the context of alleged legal malpractice, and the scholarly debate about the wisdom of such an expansion continues. Compare, for example, Deborah L. Rhode, In the Interests of Justice: Reforming the Legal Profession 166 (Oxford 2000) (describing the burden of showing quantifiable damage as a significant barrier to recovery); Lawrence W. Kessler, Alternative Liability in Litigation Malpractice Actions: Eradicating the Last Resort of Scoundrels, 37 San Diego L Rev 401, 426 (2000) (discussing high causation standards as a major failure of the legal malpractice regime), with John C.P. Goldberg, What Clients Are Owed: Cautionary Observations on Lawyers and Loss of a Chance, 52 Emory L J 1201, 1210 (2003) (“[E]ven if the argument holds for [the loss of chance doctrine to apply to] doctors in some situations, it won’t always or even usually hold for lawyers. Lawyers, in this instance, should be left less at risk of liability than doctors.”).
  • 167For simplicity’s sake here, assume full information such that we might imagine that the parties would settle a case at or around roughly its expected litigation value, if we assume away things like transaction costs, mismatching valuations, the prospect of value creation, differences in risk preferences, and a number of other real-world features of settlement.
  • 168In Scenario 1, before the lawyer’s malpractice, the plaintiff’s LV was ($1 million * 0.7) = $700,000. Following the malpractice, the LV of the case was ($1 million * 0.4) = $400,000. In Scenario 2, before the lawyer’s malpractice, the plaintiff’s LV was ($1 million * 0.4) = $400,000. Following the malpractice, the LV of the case was ($1 million * 0.1) = $100,000. In both cases, the effect was a $300,000 decrease in the LV of the case.
  • 169The plaintiff, at the moment before the lawyer’s malpractice, in our hypothetical and simplified universe, faced a choice between a certain settlement and the value of taking the case to trial—a bundle of legal endowments, the best valuation of which is captured by the LV figure.
  • 170The question is more complex with respect to circumstances in which the lawyer’s misconduct related to the valuation of a settlement option or some kind of across-the-table settlement malpractice.
  • 171This is a simplifying assumption for purposes of illustration. In the real world, the economics would be more complex, and the plaintiff would undoubtedly have a number of interests not captured by this simple way of formulating the choice. Still, the concept that there would be some point of indifference—some offer one could make to the client that would render him indifferent between accepting the offer and proceeding to trial. For purposes of this illustration, I am treating that number as the case’s LV.
  • 172This has the merits of preventing the lawyer from serving, inappropriately, as the guarantor of the case. Otherwise, the plaintiff receives all of the upside risk of proceeding to trial (where he may win outright) and it is the lawyer who bears the downside risk (because he will have to compensate the plaintiff as though the risk of taking the case to trial had paid off).
  • 173See Kiser, Beyond Right and Wrong at 23 (cited in note 70) (noting a great disparity in predictions for damages awards).
  • 174See, for example, Steven Shavell, Foundations of Economic Analysis of Law 401–11 (Belknap 2004); J.J. Prescott and Kathryn E. Spier, A Comprehensive Theory of Civil Settlement, 91 NYU L Rev 59, 71–72 (2016). The prospect also exists that settlement results from (or is prevented by) agency costs—mismatches between the attorney’s incentives and the client’s incentives. See, for example, Kathryn E. Spier and Bruce L. Hay, The Positive Theory of Litigation and Settlement in Peter Newman, ed, The New Palgrave Dictionary of Economics and the Law 442 (Stockton Press 1998) (describing the effects of fee arrangements, interests regarding publicity, timing, and control).
  • 175See, for example, Richard A. Posner, An Economic Approach to Legal Procedure and Judicial Administration, 2 J Legal Stud 399, 417–20 (1973); Steven Shavell, Suit, Settlement, and Trial: A Theoretical Analysis Under Alternative Methods for the Allocation of Legal Costs, 11 J Legal Stud 55, 63 (1982); Spier and Hay, Settlement of Litigation at 442 (cited in note 174).
  • 176There is considerable evidence that litigants develop mismatching predictions about their prospects at trial. See, for example, Kathryn E. Spier, Litigation, in A. Mitchell Polinsky and Steven Shavell, eds, 1 Handbook of Law and Economics 277–78 (North-Holland 2004) (surveying the economic literature on the effects of “[m]utual optimism,” a condition in which parties hold inconsistent views about likely trial outcomes) (emphasis in original). See generally, Spier and Hay, The Positive Theory at 443 (cited in note 174) (“The existence of divergent party expectations concerning trial remains the most influential account of why cases may fail to settle.”).
  • 177See Spier, Litigation at 269 (cited in note 176).
  • 178The literature also includes a vast array of instances in which deviations from one vision of “rationality” are described as “errors” or “mistakes.” For a thoughtful critique of this characterization, see generally Katheryn Zeiler, Mistaken about Mistakes, 48 Eur L & Econ 9 (2019).
  • 179For more on sources of settlement value, see David A. Lax and James K. Sebenius, The Manager as Negotiator: Bargaining for Cooperation and Competitive Gain 88–112 (NY Free Press 1986); Mnookin, Peppet, and Tulumello, Beyond Winning at 225–26 (cited in note 4); Leigh L. Thompson, The Mind and Heart of the Negotiator 76–87 (Pearson 2012). See also McDermott, Inc v AmClyde, 511 US 202, 215 (1994) (“The parties’ desire to avoid litigation costs, to reduce uncertainty, and to maintain ongoing commercial relationships is sufficient to ensure nontrial dispositions in the vast majority of cases.”).
  • 180See, for example, Thompson, The Mind and Heart of the Negotiator at 74 (cited in note 179) (“Most untrained negotiators view negotiation as a pie-slicing task: They assume that their interests are incompatible, that impasse is likely, and that issues are settled one by one rather than as packages.”); Kathleen M. O’Connor and Ann A. Adams, What Novices Think about Negotiation: A Content Analysis of Scripts, 15 Negotiation J 135, 142 (1999) (“[N]ovices tend to assume that negotiator interests are incompatible.”).
  • 181See, for example, French v Domnarski, 1995 WL 573787, *1 (Conn Super) (describing a client who settled a claim, relying on an attorney’s mistaken advice that alimony and certain mortgage payments would be tax deductible); Jalali v Root, 109 Cal App 4th 1768, 1775 (2003) (explaining that a client relied on an attorney’s allegedly erroneous advice about the tax consequences of payments when deciding to settle sexual harassment claim).
  • 182For an example, see Mnookin, Peppet, and Tulumello, Beyond Winning at 240–47 (cited in note 4) (describing a deal between Digital and Intel resolving a patent dispute through the use of a series of business transactions).
  • 183See, for example, Draper v Brennan, 713 A2d 373, 374 (NH 1998) (defendant in employment dispute agreed to keep former employee plaintiff on company insurance program until he reached age sixty-five, but plaintiff’s attorney failed to secure protection against the company later charging plaintiff for these benefits).
  • 184See, for example, Steinberg v Grasso, 2007 WL 701689, *2 (NJ Super) (describing a lawyer who incorrectly advised a client that a proposed agreement contained a high-low partial agreement that would shield the client from having to report a claim to the national practitioner data bank). But see Zalta v Billips, 81 Cal App 3d 183, 188 (1978) (describing plaintiffs’ claim that their lawyer failed to correct a mistaken summary of a settlement, causing them reputational damage when the misreported settlement suggested that the plaintiffs were admitting medical malpractice); Barella v Exchange Bank, 84 Cal App 4th 793, 801 (2000) (“[T]he value to a particular plaintiff of public vindication (or, conversely, the negative value of confidentiality) is so highly subjective and elusive that no court can determine its monetary worth.”). This tendency of courts to conflate parties’ interests with the legal remedies articulated in pleadings is widespread but troublesome. See Moffitt, 80 Ind L Rev at 744–45 (cited in note 10). We know, however, that real-world clients “may seek compensation, vengeance, fair and dignified treatment, apologies, reform, and an array of other goals in their legal interactions.” Robbennolt and Sternlight, Psychology for Lawyers at 258 (cited in note 39).
  • 185See, for example, Mnookin, Peppet, and Tulumello, Beyond Winning at 27–43 (cited in note 4); Thompson, The Mind and Heart of the Negotiator at 40–68 (cited in note 179).
  • 186457 SW3d 765 (Mo App 2014).
  • 187Id at 767. In the legal malpractice suit that followed, the court rejected as “speculative” the plaintiffs’ assertion that this conduct prejudiced them in subsequent negotiations with Wal-Mart. Id at 769.
  • 188See, for example, FRE 408; Eisenmann v Podhorn, 528 SW3d 22 (Mo App 2017) (“Evidence of settlement offers or agreements is generally inadmissible because public policy favors the settlement of disputes, and because offers of settlement ‘lend an aura of guilt and/or liability to the offering party.’”), citing Ullrich v CADCO, Inc, 244 SW3d 772, 780 (Mo App 2008).
  • 189One’s reservation value is typically a function of one’s BATNA, and there are contexts in which revealing one’s BATNA may yield strategic benefit. In simplistic terms, if I believe your walkaway alternative is lousy, I may have insufficient incentive to put something attractive on the table in our negotiations. You may, therefore, perceive the need to share your perception of your BATNA with me, in hopes that I will be persuaded that you will walk away if I am not more generous. Whether one ought to disclose a reservation value is a more debatable proposition. The easiest case of settlement malpractice, therefore, is one in which the disclosure was unauthorized and accidental, rather than a function of an intentional negotiation strategy.
  • 190Note that this may require courts to reduce noneconomic interests into economic terms, given their remedial limitations. This is awkward, to be certain, but it must still be better than simply assuming those real-world interests away. See Moffitt, 80 Ind L J at 744–47 (cited in note 10) (describing ways in which pleadings provide inadequate articulations of clients’ true interests).
  • 191The Supreme Court has recognized, in at least limited criminal contexts, the constitutional importance of client choice. See, for example, McCoy v Louisiana, 138 S Ct 1500, 1507 (2018) (explaining that a defendant’s choices in exercising the right to defend rather than plead guilty must be respected). But this recognition has not extended to include a calculable measure of damages.
  • 192834 F2d 1105 (1st Cir 1987).
  • 193Id at 1106.
  • 194Id at 1107–08.
  • 195Id at 1107.
  • 196Moores, 834 F2d at 1110.
  • 197Id at 1109–13.
  • 198Model Rules, Rule 1.2(a).
  • 199Model Rules, Rule 1.4(b).
  • 200See, for example, Commissioner v Banks, 543 US 426, 436 (2005); Ex parte Free, 910 S2d 753, 755, 757 (Ala 2005) (reversing a trial court’s dismissal of a claim in which the plaintiff alleged that her attorney presented the disabled plaintiff “with merely 4 signature pages, simply advising her to sign them because ‘the Judge is in a hurry and there isn’t enough time for me to read them to you’”).
  • 201See, for example, In re Plaza, 363 Bankr 517, 520, 522 (Bankr SD Tex 2007) (refusing to enforce a provision providing that the “client will not make settlement . . . or accept any sum as reimbursement for any of the client’s injuries or expenses, without the attorney’s consent,” on the grounds that it “appears to be directly at odds with the well accepted principle that it is the client who has exclusive control over whether to settle, compromise or adjust the cause of action”).
  • 202I have no easy answer to the question of what value to assign to client choice, in which the existing framework provides for ready comparative calculation resulting in an award. Perhaps in some contexts the deal should be more readily voidable than might otherwise be permitted under the existing system. Perhaps the lawyer should become the guarantor of the uncommunicated settlement offer. Perhaps some part of the lawyer’s fees should be forfeit as an imperfect proxy for the value the client attaches to choice. Or perhaps, the “loss of chance” doctrine should apply in the legal malpractice context.
  • 203See, for example, William F. Coyne Jr, The Case for Settlement Counsel, 14 Ohio St J Disp Resol 367, 392 (1999); James McGuire, Why Litigators Should Use Settlement Counsel, 18 Alternatives to High Cost of Litig 119, 120 (2000). See also generally Kathy A. Bryan, Why Should Businesses Hire Settlement Counsel?, 2008 J Disp Resol 195.
  • 204Coyne, 14 Ohio St J Disp Resol at 369 (cited in note 203):

    [T]here are significant incentives for lawyers not to embrace early settlement. These incentives include the need to market services, the desire not to appear weak, the obligation to represent a client zealously, the thirst for justice, and last, but perhaps not least, the desire to maximize income. In addition, it is extremely difficult, psychologically, for an attorney to act as an effective advocate and, at the same time, to encourage settlement.

  • 205See generally Pauline Tesler, Collaborative Law: Achieving Effective Resolution in Divorce without Litigation (ABA 2001); William H. Schwab, Collaborative Lawyering: A Closer Look at an Emerging Practice, 4 Pepperdine Disp Resol L J 351 (2004). For a foreshadowing of the Collaborative Law experiment, see Ronald J. Gilson and Robert H. Mnookin, Disputing through Agents: Cooperation and Conflict between Lawyers in Litigation, 94 Colum L Rev 509, 513, 522–34 (1994) (“[L]awyers are repeat players who have the opportunity to establish reputations. At the core of our story is the potential for disputing parties to avoid the prisoner’s dilemma inherent in much litigation by selecting cooperative lawyers whose reputations credibly commit each party to a cooperative strategy.”).
  • 206See Schwab, 4 Pepperdine Disp Resol L J at 358–60 (cited in note 205).
  • 207For more on the signaling aspects of collaborative law, see Scott R. Peppet, Lawyers’ Bargaining Ethics, Contract, and Collaboration: The End of the Legal Profession and the Beginning of Professional Pluralism, 90 Iowa L Rev 475, 481–97 (2005).
  • 208See Scott R. Peppet, The Ethics of Collaborative Law, 2008 J Disp Resol 131, 133.
  • 209Collaborative law still presents the risk of self-dealing if, in order to continue to amass hourly fees, the lawyer recommends continued negotiations.
  • 210Most insurers have reported significantly lower risk premiums for lawyers who work in practice areas in which mediation and negotiation are prominent. See Tom Baker and Rick Swedloff, Liability Insurer Data as a Window on Lawyers’ Professional Liability, 5 UC Irvine L Rev 1273, 1291 n 60 (2015). It would be interesting to see if any differentiation emerges in the marketplace for insurance.
  • 211For a useful overview, see generally Susan M. Treyz, Criminal Malpractice: Privilege of the Innocent Plaintiff, 59 Fordham L Rev 719 (1991).
  • 212See Gideon v Wainwright, 372 US 335, 351 (1963) (recognizing the right to counsel for offenses carrying the possibility of substantial prison sentences); McMann v Richardson, 397 US 759, 771 n 14 (1970) (“It has long been recognized that the right to counsel is the right to the effective assistance of counsel.”); Strickland v Washington, 466 US 668, 687 (1984) (establishing the relevant two-part standard for determining whether representation was ineffective).
  • 213559 US 356 (2015).
  • 214Id at 359.
  • 215Jenny Roberts, Effective Plea Bargaining Counsel, 122 Yale L J 2650, 2657 (2013). Note the parallels with Ordon. See Part III.A.
  • 216Padilla, 559 US at 359.
  • 217Frye, 556 US at 39.
  • 218Id at 138–39. Note the parallels with Moores. See Part III.C.3.
  • 219566 US 156 (2012).
  • 220Id at 161.
  • 221Id. Note the parallels with Sauer. See Part III.A.
  • 222Padilla, 559 US at 374; Frye, 566 US at 151 (remanding to consider state law issues relevant for the second prong under Strickland); Lafler, 566 US at 174.
  • 223Strickland, 466 US at 687. See also Part II.B.
  • 224As the Court in Frye wrote:

    “The art of negotiation is at least as nuanced as the art of trial advocacy and it presents questions farther removed from immediate judicial supervision.” . . . Bargaining is, by its nature, defined to a substantial degree by personal style. The alternative courses and tactics in negotiation are so individual that it may be neither prudent nor practicable to try to elaborate or define detailed standards for the proper discharge of defense counsel’s participation in the process.

    Frye, 566 US at 144–45, quoting Premo v Moore, 562 US 115, 125 (2011). See also Part II.A.

  • 225See Roberts, 122 Yale L J at 2671–72 (cited in note 215). See also Part III.B.
  • 226See, for example, Rishi Batra, Judicial Participation in Plea Bargaining: A Dispute Resolution Perspective, 76 Ohio St L J 565, 589–90 (2015); Nancy J. King and Ronald F. Wright, The Invisible Revolution in Plea Bargaining: Managerial Judging and Judicial Participation in Negotiations, 95 Tex L Rev 325, 347–56 (2016). For an illustration of the limits and potential complications of judicial involvement—even in the sense of a judicial colloquy—on the prospects of a subsequent malpractice claim, see Puder v Buechel, 874 A2d 534, 538 (NJ 2005) (evaluating whether continued prosecution of a legal malpractice claim would violate principles of judicial estoppel due to plaintiff’s representations to the court that the prosecuted claims were moot).
  • 227The common law arrow may also sometimes flow the other direction—that is, that developments in the realm of civil settlement malpractice may inform courts’ treatment of alleged plea negotiation malpractice.
  • 228See Carrie Menkel-Meadow, Mothers and Fathers of Invention: The Intellectual Founders of ADR, 16 Ohio St J Disp Resol 1, 1–5 (2000) (providing historical perspective on the field’s development).
  • 229Online dating services, policing focused on crime “hot spots,” calculating routes for delivery drivers, and the “Moneyball” approach to managing a professional sports team are conspicuous examples. Steve Lohr, The Age of Big Data (NY Times, Feb 11, 2012), archived at See also Taha Yasseri, How Big Data Will Change Our Lives and Our Understanding of Them (Dataconomy, May 16, 2014), archived at
  • 230Such valuation guidance would be feasible only with access to greater information about settlements and their terms. We have seen movements toward making at least some kinds of settlements more public. For example, the “Harvey Weinstein tax reform” now provides a disincentive for confidentiality and nondisclosure clauses in certain kinds of settlements. See 26 USC § 162(q). Similarly, some state laws refuse to enforce, on public policy grounds, confidentiality clauses in settlements affecting public health. See, for example, Jillian Smith, Secret Settlements: What You Don’t Know Can Kill You!, 2004 Mich St L Rev 237; Erik S. Knutsen, Keeping Settlements Secret, 37 Fla St U L Rev 945 (2010); Saul Levmore and Frank Fagan, Semi-Confidential Settlements in Civil, Criminal, and Sexual Assault Cases, 103 Cornell L Rev 311 (2018).
  • 231But see Scott R. Peppet and Michael L. Moffitt, Learning How to Learn to Negotiate in Chris Honeyman and Andrea Kupfer Schneider, eds, The Negotiator’s Desk Reference 13 (DRI 2017) (describing concrete data, such as that available in a transcript, as critical to action-science-based learning).
  • 232The meteoric rise of online dispute resolution makes access to data at this scale a distinct possibility. See, for example, Elayne E. Greenberg and Noam Ebner, What Dinosaurs Can Teach Lawyers about How to Avoid Extinction in the ODR Evolution *10–11 (St. John’s School of Law Legal Studies Research Paper Series, Jan 17, 2019), archived at; Scott J. Shackleford and Anjanette H. Raymond, Building the Virtual Courthouse: Ethical Considerations for Design, Implementation, and Regulation in the World of ODR, 2014 Wis L Rev 615, 622. See also generally Ethan Katsch and Colin Rule, What We Know and Need to Know about Online Dispute Resolution, 67 SC L Rev 329 (2016).
  • 233The dearth of information about lawyers’ conduct in negotiations is emblematic of the lack of information about lawyers’ conduct more generally. See Randall Kiser, Beyond Right and Wrong at 11 (cited in note 70) (describing what he calls “The Paradox of Copious Lawyers and Scant Data”).
  • 234With apologies to every English teacher I have ever known, I acknowledge that single sentence exceeds the number of negatives typically permitted in an entire paragraph. Perhaps there would be a way to word this sentence differently, but I think it is important to highlight the difference in prospects between looking for what is effective and looking for what might be ineffective.
  • 235Hadfield and Rhode, 67 Hastings L J at 1201 (emphasis in original) (cited in note 38).
  • 236See Michael Moffitt, Islands, Vitamins, Salt, Germs: Four Visions of the Future of ADR in Law Schools (and a Data-Driven Snapshot of the Field Today), 25 Ohio St J Disp Resol 25, 31 (2010).
  • 237ABA Directory (ABA and University of Oregon School of Law), archived at
  • 238Compare Section of Dispute Resolution (ABA), archived at (stating that there are 10,000 members), with Section of Family Law (ABA), archived at (stating that there are 12,000 members), and Health Law Section (ABA), archived at (stating that there are 10,600 members).
  • 239See, for example, The Harvard Negotiation Law Review; The Ohio State Journal on Dispute Resolution; Journal of Dispute Resolution (University of Missouri School of Law); Pepperdine Dispute Resolution Law Journal; Willamette Journal of International Law and Dispute Resolution; Cardozo Journal of Conflict Resolution.
  • 240See, for example, Dispute Resolution (Certificate) (University of Missouri School of Law 2019), archived at; Dispute Resolution Certificate Program (Pepperdine Law 2019), archived at; Concentrations: Dispute Resolution (Oregon School of Law), archived at
  • 241See Dispute Resolution Syllabi (The University of Missouri, 2019), archived at (collecting syllabi from a wide variety dispute resolution courses).
  • 242The only hesitation I have in making this categorical assertion is that the principal text used in two of the available syllabi does not include the use of the acronym “BATNA.” It does, however, urge students to consider “nonsettlement alternatives.” Charles Craver, Effective Legal Negotiation and Settlement 142 (Carolina 2016).
  • 243See, for example, Michelle M. Mello and Troyen A. Brennan, Deterrence of Medical Errors: Theory and Evidence for Malpractice Reform, 80 Tex L Rev 1595, 1615 (2002) (finding only “limited evidence” supporting the theory that tort liability serves as a deterrent); Gary T. Schwartz, Reality in the Economic Analysis of Tort Law: Does Tort Law Really Deter?, 42 UCLA L Rev 377, 422–30 (1994) (suggesting tort liability produces, at most, a “weak” form of deterrence); Lydia Nussbaum, Trial and Error: Legislating ADR for Medical Malpractice Reform, 76 Md L Rev 247, 257–60 (2017) (surveying research on the mismatch between medical negligence and patient recovery in medical malpractice cases).
  • 244Wilkins, 105 Harv L Rev at 831 (cited in note 25). See also Goldberg, 59 Vand L Rev at 1079 (cited in note 38) (“Perhaps state governments or the federal government ought to put in place systems of regulation and compensation that operate apart from the tort system, or schemes that foster conditions that will permit market forces to generate incentives toward safety.”).
  • 245For a thoughtful argument that the bar ought to apply restorative justice principles in the context of attorney discipline, see generally Jennifer Gerarda Brown and Liana G.T. Wolf, The Paradox and Promise of Restorative Attorney Discipline, 12 Nev L J 253 (2012).
  • 246Perhaps we will see a greater role for judicial colloquies in the civil context, similar to the role they play in the criminal context. See, for example, Julian A. Cook III, Crumbs from the Master’s Table: The Supreme Court, Pro Se Defendants and the Federal Guilty Plea Process, 81 Notre Dame L Rev 1895, 1900 (2006); J. Vincent Aprile II, Waiving the Integrity of the Criminal Justice System, 24 Crim Just 46, 51 (2010). See also Puder, 874 A2d at 536–38. Or perhaps judicial involvement in settlement conferences will increase in ways that provide some measure of oversight. See Carrie Menkel-Meadow, For and Against Settlement: Uses and Abuses of the Mandatory Settlement Conference, 33 UCLA L Rev 485, 506–07 (1985); Ellen E. Deason, Beyond “Managerial Judges”: Appropriate Roles in Settlement, 78 Ohio St L J 73, 73, 105 (2017) (proposing structural limitations on judges serving dual-neutral roles of “managing a settlement process” and “serving as a settlement neutral”). Each of these would, of course, be complicated.
  • 247See, for example, John C.P. Goldberg and Benjamin C. Zipursky, From Riggs v. Palmer to Shelley v. Kraemer: The Continuing Significance of the Law-Equity Distinction *7–9 (Harvard Public Law Working Paper, Nov 7, 2018), archived at
  • 248Moffitt, 80 Ind L Rev at 728 (cited in note 10).
  • 249See Part I.
  • 250For example, I chose not to study Pennsylvania, because the Muhammad rule precludes a number of malpractice actions that would not be barred in other jurisdictions. In this way, the data may slightly overstate the frequency with which settlement malpractice claims arise as a nationwide matter. See generally Muhammad, 587 A2d 1346. See also Part II.B.
  • 251See, for example, McVeigh v Fleming, 410 SW3d 287, 288 (Mo App 2013) (dispute regarding whether a client must contribute to his attorney’s photocopying costs); Tankersley v Almand, 73 F Supp 3d 629, 631–33 (D Md 2014) (considering whether a lawyer was required to provide his social security number to keep his license).
  • 252See generally, for example, Quinn v Fishkin, 117 F Supp 3d 134 (D Conn 2015); Beaudry v Harding, 104 A3d 134 (Me 2014).
  • 253Representatives from both Westlaw and Lexis suggested almost identical Boolean search terms. See Online Interview with LexisNexis Support Representative (July 18, 2018) (on file with author). See also Online Interview with Westlaw Reference Attorney (July 17, 2018) (on file with author).
  • 254See, for example, Kaye v Wilson-Gaskins, 135 A3d 892, 903 (Md App 2016); Spitz v St. Luke’s Medical Center, 2007 WL 926391 *2 (Ohio App); Wiegand v Wiegand, 21 A3d 489, 491 (Conn App 2011); Bridge v Phoenix Bond & Indemnity Co, 553 US 639, 650 (2008).
  • 255See, for example, Oliver v National Collegiate Athletic Association, 920 NE2d 203, 215 (Ohio Com Pl 2009) (negotiating sports contracts for college athletes).
  • 256Gray v Weinstein, 955 A2d 1246, 1253 (Conn App 2008).
  • 257Barrow v Walsh, 2011 WL 4716283, *3 (Conn Super).
  • 258Disciplinary Counsel v Schmidt, 983 NE2d 1310, 1313 (Ohio 2012).
  • 259Joyce v Rough, 2008 WL 4763485, *1 (Ohio App).
  • 260Stender v Blessum, 897 NW2d 491, 507 (Iowa 2017), quoting Kling v Landry, 686 NE2d 33, 40 (Ill App 1997). See also Jerman v Carlisle, McNellie, Rini, Kramer & Ulrich LPA, 559 US 573, 612 (2010) (“[T]he Court, by failing to adopt a reasonable interpretation to counter these excesses, risks compromising its own institutional responsibility to ensure a workable and just litigation system.”).
  • 261In re Application of Bell, 861 NE2d 533, 534 (Ohio 2007).
  • 262Wisconsin is the only state with an online database of disciplinary opinions formatted in a way that permits substantive Boolean searches with date restrictors. It did not, however, permit a global search about the total number of cases, so I omitted it from my list.
  • 263See, for example, 2017 Joint Annual Report: State of Michigan Attorney Grievance Commission Annual Report (State of Michigan Attorney Discipline Board and Attorney Grievance Commission, 2017), archived at; Lawyer Discipline Statistics (Florida Bar, 2017), archived at
  • 264See, for example, Lawyer Discipline Statistics (Florida Bar, 2017), archived at (breaking down complaints by type of reprimand). See also Index to Ethics Opinions (North Carolina State Bar), archived at (categorizing by practice area).
  • 265The ABA curates the National Lawyer Disciplinary Data Bank, but it permits only searches for particular lawyers, not for particular disciplinary actions across the population of lawyers. Its use is principally for “disciplinary authorities and bar admissions agencies in providing a central repository of information to facilitate reciprocal discipline and to help prevent the admission of lawyers who have been disbarred or suspended elsewhere.” National Lawyer Regulatory Data Bank (ABA, Sept 10, 2018), archived at