Most everyone agrees that in the American civil justice system many important legal rights go unvindicated, serious losses remain uncompensated, and those called on to defend their conduct are often forced to spend altogether too much. Eighty percent of the members of the American College of Trial Lawyers report that pretrial costs and delays keep injured parties from bringing valid claims to court. Seventy percent also say attorneys use the threat of discovery and other pretrial costs as a means to force settlements that aren’t based on the merits.
The upshot? Legal services in this country are so expensive that the United States ranks near the bottom of developed nations when it comes to access to counsel in civil cases.
The real question is what to do about it.
This paper explores three possible avenues for reform. All three lie within the power of the legal profession to effect. They include revisions to our ethical codes, civil justice rules, and legal education accreditation requirements — possibilities that in turn challenge each of the main elements of our profession: bar, bench, and academy. Each of these avenues of reform holds the promise of either reducing the cost or increasing the output of legal services — in that way making access to justice more affordable. And for that reason, you might think of them as (sort of) market-based solutions.
Now, you might wonder why this paper doesn’t address some other angles at change — perhaps most obviously the possibility of increased public financing for legal aid. One reason is that, whatever challenges may be associated with asking a self-regulating profession to reconsider its self-imposed barriers to entry and output restrictions, entering that political and fiscal thicket appears likely to pose even more. Maybe even more importantly, though, on the road to change perhaps we should begin by asking first what we can do on our own and without expense to the public fisc, and whether and to what degree our own self-imposed rules increase the cost of legal services and decrease access to justice in unwarranted ways?
We lawyers enjoy a rare privilege. We are largely left to regulate our own market, often through rules of our own creation and sometimes through statutes effectively of our own devise. Of course and no matter the industry, even the most well-intentioned regulations can bear negative unintended consequences. Sometimes even the intended consequences of regulations can only be described as rent-seeking. And it seems hard to think our profession might be immune from these risks. Surely many of our self-imposed regulations represent well-intentioned efforts to prevent and police misconduct that risks harm to clients. But you might also wonder if a profession entrusted with the privilege of self-regulation is at least as (or maybe more) susceptible than other lines of commerce to regulations that impose too many social costs compared to their attendant benefits. Consider two examples.
Unauthorized Practice of Law
Marcus Arnold presented himself as a legal expert on AskMe.com, a website that allows anyone to volunteer answers to posted questions. Users of the site rate those who offer advice, and in time they came to rank Arnold as the third most helpful volunteer of legal answers out of about 150 self-identified legal experts. When Arnold later revealed that he was but a high school student, howls emerged from many quarters and his ranking dropped precipitously. Still, his answers apparently continued to satisfy the website’s users because soon enough he went on to attain the number one ranking for legal advice, ahead of scores of lawyers. Like a Rorschach test, both supporters and opponents of unauthorized practice of law (UPL) regulations see in this case support for their positions.
When approaching questions about the unauthorized practice of law, you might think it’s a natural place to begin by asking what exactly constitutes the practice of law. But that turns out to be a pretty vexing little question. While the ABA offers a set of model rules of professional conduct governing those who engage in the practice of law, it is surely a curiosity that those rules don’t attempt to define what constitutes the practice of law in the first place. After all, it’s no easy thing to regulate an activity without first defining what that activity is.
The fact is the job of defining what does and doesn’t constitute the practice of law has largely been left to state statutes. And history reveals that the definitions states have adopted, usually at the behest of local bar associations, are often breathtakingly broad and opaque — describing the practice of law as, and prohibiting nonlawyers from participating in, the “represent[ation]” of others, or (even more circularly) any “activity which has traditionally been performed exclusively by persons authorized to practice law.” More than a few thoughtful people have wondered if these sorts of sweeping and opaque restrictions may be subject to constitutional challenge on vagueness, First Amendment, or due process grounds.
But however that may be, about one thing there can be little doubt. In recent years, lawyers have used the expansive UPL rules they’ve sought and won to combat competition from outsiders seeking to provide routine but arguably “legal” services at low or no cost to consumers. Indeed, by far and away most UPL complaints come from lawyers rather than clients and involve no specific claims of injury. Take recent cases involving Quicken Family Lawyer and LegalZoom. Those firms sell software with forms for wills, leases, premarital agreements, and dozens of other common situations. When Quicken entered the Texas market, an “unauthorized practice of law committee” appointed by the Texas Supreme Court quickly brought suit, a fight that eventually yielded a federal court decision holding that Quicken had violated Texas UPL regulations (though, happily, a result the legislature later effectively undid). Similarly, when LegalZoom entered the market in North Carolina, the state bar declared its operations illegal, a declaration that eventually induced the company to settle and promise to revise some of its business practices. Neither are challenges of this sort aimed only at for-profit firms. The federal Individuals with Disabilities Education Act (IDEA) affords parents the right to be “accompanied and advised” in agency proceedings by nonlawyers who have special training or knowledge “with respect to the problems of children with disabilities.” Yet even here, where (supreme?) federal law seems clear, state authorities have sought (sometimes successfully) to use UPL laws to forbid lay advocacy by nonprofit firms with expertise in IDEA procedures. To be sure, efforts like these to thwart competition from commercial and nonprofit advocates have proven only partially successful — LegalZoom and companies like it continue to expand. But surely, too, the threat and costs of litigation deter entry by others and raise costs for those who do enter, costs the consumer must ultimately bear.
It seems well past time to reconsider our sweeping UPL prohibitions. The fact is nonlawyers already perform — and have long performed — many kinds of work traditionally and simultaneously performed by lawyers. Nonlawyers prepare tax returns and give tax advice. They regularly negotiate with and argue cases before the Internal Revenue Service. They prepare patent applications and otherwise advocate on behalf of inventors before the Patent & Trademark Office. And it is entirely unclear why exceptions should exist to help these sort of niche (and some might say, financially capable) populations but not be expanded in ways more consciously aimed at serving larger numbers of lower- and middle-class clients.
Some states are currently experimenting with intriguing possibilities. California now licenses “legal document assistants” who may help consumers before certain tribunals. Colorado permits nonlawyers to represent claimants in unemployment proceedings. And Washington allows legal technicians to assist clients in domestic relations cases provided they meet certain requirements — like obtaining an associate’s degree, passing an exam, completing 3,000 hours of supervised paralegal work, and taking certain legal courses. The ABA itself recently partnered with one of LegalZoom’s competitors, Rocket Lawyer, to help the association’s members connect with potential clients online, in the process seemingly granting its imprimatur to a company that some argue engages in the unauthorized practice of law.
Consistent with the law of supply and demand, increasing the supply of legal services can be expected to lower prices, drive efficiency, and improve consumer satisfaction. And, in fact, studies suggest that lay specialists who provide representation in bankruptcy and administrative proceedings often perform as well as or even better than attorneys and generate greater consumer satisfaction. The American Law Institute has noted, too, that “experience in several states with extensive nonlawyer provision of traditional legal services indicates no significant risk of harm to consumers.” And the Federal Trade Commission has observed that it is “not aware of any evidence of consumer harm arising from [the provision of legal services by nonlawyers] that would justify foreclosing competition.” In the United Kingdom, where nonlawyers can win government contracts to provide legal advice and appear before some administrative tribunals, nonlawyers significantly outperform lawyers in terms of results and satisfaction when dealing with low-income clients. Indeed, studies there show that the best predictor of quality appears to be “specialization, not professional status.”
Of course, the potential for abuse cannot be disregarded. Many thoughtful commentators suggest that UPL restrictions are necessary to protect the public from fraudulent or unqualified practitioners. And surely many lay persons, and perhaps most especially the most underserved, are not well equipped to judge legal expertise. But do these entirely valid concerns justify the absolute UPL bans found today in so many states? That seems an increasingly hard case to make in light of an increasing amount of evidence suggesting that, at least in specified practice areas, a more nuanced approach might adequately preserve (or even enhance) quality while simultaneously increasing access to competent and affordable legal services.
All else equal, market participants with greater access to capital can increase output and lower price. So, for example, optometry, dental, and tax preparation services are no doubt cheaper and more ubiquitous today thanks to the infusion of capital from investors outside those professions. Indeed, consumers can often now find all these services (and more) in their local “superstores.” Yet Rule 5.4 of the ABA’s Model Rules of Professional Conduct — adopted by most states — prohibits nonlawyers from obtaining “any interest” in a law firm. So while consumers may obtain basic medical and accounting services cheaply and conveniently in and thanks to (say) Walmart, they can’t secure similar assistance with a will or a landlord-tenant problem. With a restricted capital base (limited to equity and debt of individual partners), the output of legal services is restricted and the price raised above competitive levels, for as Prof. Stephen Gillers has put it, “lay investors might be willing to accept a lower return on their money” than lawyers shielded by Rule 5.4.
Rule 5.4 bears a curious history. After thoroughly studying the issue, the commission that created the first draft of the model rules back in 1982 suggested that lawyers should be allowed to work in firms owned or managed by nonlawyers. But this suggestion was defeated in the ABA House of Delegates and replaced by the present rule effectively preventing nonlawyers from aquiring “any interest” in a law firm. Since then, ABA committees have repeatedly proposed changes to Rule 5.4 but every proposal has, like the first, gone down to defeat in the House of Delegates. Most recently, in 2009 an ABA commission supported serious consideration of three alternatives to the rule. The most modest option would have (1) required a firm to engage only in the practice of law, (2) prohibited nonlawyers from owning more than a certain percentage (e.g., 25 percent) of a firm, and (3) demanded that nonlawyer owners pass a “fit to own” test. Another approach would have allowed lawyers to engage in partnerships of this sort without the cap on nonlawyer ownership or the fit to own test. And the third and final option would have done away with all three requirements and permitted firms to offer both legal and nonlegal services.
Notably, the United Kingdom has permitted multidisciplinary firms and nonlawyer investment since 2007. In the first two years of the program, 386 so-called “alternative business structures” (ABSs) were established. Six years into the experiment, the Solicitors Regulatory Authority analyzed ABSs and found that while these entities accounted for only 3 percent of all law firms, they had captured 20 percent of consumer and mental health work and nearly 33 percent of the personal injury market — suggesting that ABSs were indeed serving the needs of the poor and middle class, not just or even primarily the wealthy. Notably, too, almost one-third of ABSs were new participants in the legal services market, thus increasing supply and presumably decreasing price. ABSs also reached customers online at far greater rates than traditional firms — over 90 percent of ABSs were found to possess an online presence versus roughly 50 percent of traditional firms, again suggesting an increased focus on reaching individual consumers. Given the success of this program, it’s no surprise that some U.S. jurisdictions have appointed committees to study reforms along just these lines.
Of course, supporters of the current ABA ban contend that allowing nonlawyers to participate in legal practice might influence lawyers’ professional judgment. But it is again worth asking whether these entirely legitimate concerns justify a total ban on the practice. After all, we routinely address similar independence concerns in the model rules without resort to total bans. So, for example, we permit third parties (e.g., insurance companies) to pay for an insured’s legal services but restrict their ability to interfere with the attorney-client relationship. We allow in-house counsel to work for corporations where they must answer to executives but require them sometimes to make noisy withdrawals. And we increasingly permit law firms to manage client and personal financial conflicts by screening affected lawyers rather than by banning the firm from representing a client. Of course, in each of these cases lawyers stand to benefit from rules that permit an engagement that might otherwise be forbidden. But surely it shouldn’t be the case that we will forgo or lift outright bans in favor of more carefully tailored rules only when we stand to gain.
The Federal Rules of Civil Procedure aim to shepherd parties toward “the just, speedy, and inexpensive determination of every action and proceeding.” But as the American College of Trial Lawyers’ survey suggests, it seems the rules sometimes yield more nearly the opposite of their intended result: expensive and painfully slow litigation that is itself a form of injustice. After years of study, the federal rules committees recently advanced a package of amendments (the “Duke Package”) seeking to address the problem. The Duke Package made three important changes. It emphasized proportionality as the governing principle for discovery. It tightened discovery deadlines and so shortened the opportunities for delay. And it sought to reduce costs by increasing certainty about parties’ obligations to preserve electronically stored information.
While these changes are no doubt a start, it’s hard to imagine they’ll finish the job of realizing the promise of Rule 1 in the 21st century. After all, our so-called “modern” rules of civil procedure are now almost 80 years old, written for an age in which discovery involved the exchange of mimeographs, not metadata. Neither do you have to look far to see promising models of change. In recent years, at least 30 states and federal district courts have implemented pilot projects testing various amendments to our long-in-the-tooth rules, all with an eye on increasing the efficiency and fairness of civil justice administration. Not every project has proven a resounding success, but the results suggest at least two other possible avenues for reform, and the federal rules committees are contemplating pilot projects to test both in the federal system.
Early and Firm Trial Dates
A RAND study of the federal judicial system in the 1990s found (perhaps to no litigator’s surprise) that setting a firm and early trial date is the single “most important” thing a court can do to reduce time to disposition. A more recent IAALS study found the same thing: a strong positive correlation between time to resolution and the elapsed time between the filing of a case and the court’s setting of a trial date. Studies of recent experiments in Oregon, Colorado, and other state court systems have shown, as well, that firm and early trial dates contribute to reducing litigation costs and increasing client and lawyer satisfaction. And in light of so much data like this, IAALS, the College, and the National Conference of Chief Justices have all recently endorsed the setting of an early and firm trial date as a best practice in civil litigation. Yet, despite this mounting evidence, and while some federal districts today adhere to the practice of setting a firm and early trial date in every case (e.g., the Eastern District of Virginia), systemwide in our federal courts over 92 percent of motions to continue trial dates are granted and fewer than 45 percent of cases that go to trial do so on the date originally set by the court.
Naturally, the possibility of mandating the practice of setting early and firm trial dates will raise some legitimate concerns. Like the worry that reducing time for trial preparation may not afford complicated cases the time and attention they require. Or the worry that deadlines set early in a case may prove too rigid to account for developments that arise only later. No doubt concerns like these suggest the importance of accounting for a case’s complexity when setting a trial date (perhaps examining empirical data regarding how long certain classes of cases take to prepare would be helpful here, data the federal courts now collect and share with judges routinely). Concerns like these may suggest as well the need to preserve a measure of flexibility to respond to new developments — perhaps by permitting continuances in “extraordinary circumstances.” But just as important is what concerns like these don’t suggest: reason to ignore the proven empirical benefits of setting an (appropriately) early and (normally quite) firm trial date in every single case.
In 1993, the federal rules committees experimented with a rule requiring parties to disclose evidence and documents both helpful and harmful to their respective causes at the outset of discovery. As the committees reasoned, lawyers and parties are rightly expected to fight over the merits but that doesn’t necessarily mean they should be permitted to fight (sometimes seemingly endless) collateral battles over what facts they must share with the other side. Just as a prosecutor must reveal exculpatory Brady material before proceeding to a vigorous fight on the merits, so too civil parties should have to disclose the good and the bad of their evidence before proceeding to litigate its significance.
The proposal met with swift criticism. Some argued that requiring lawyers to produce discovery harmful to their clients asks them to violate their clients’ trust. Others questioned whether a lawyer for one side is well positioned to know what might be helpful to the other. In response to criticisms like these, the rules committees permitted districts to opt out of the initial disclosure requirement, and a number did so, resulting in a patchwork of practices nationwide. And then, responding to complaints about this development, the committees in 2000 narrowed the mandatory-disclosure rule to require only the production of helpful evidence.
That might have seemed the end of it. Except that since 2000 a number of states have returned to the idea of mandating early and broad disclosures. And in that time a good deal of evidence has emerged suggesting these disclosures allow parties to focus more quickly and cheaply on the merits of their litigation. For example, Arizona requires parties to disclose all documents they believe to be “relevant to the subject matter of the action” within 40 days after a responsive pleading is filed. In 2009, an IAALS survey found Arizona litigators preferred state to federal court practice on this score by a 2-to-1 margin. Respondents confirmed that Arizona’s rule “reveal[s] the pertinent facts early in the case” (76 percent), “help[s] narrow the issues early” on (70 percent), and facilitates agreement on the scope and timing of discovery (54 percent). Similarly, respondents disagreed with the notion that the disclosure rule either adds to the cost of litigation (58 percent) or unduly frontloads investment in a case (71 percent). Importantly, too, counsel for plaintiffs and defendants responded in largely the same way on all these issues.
Other states and even a recent experiment in the federal system have reported similar results. A pilot project in Colorado requiring robust early disclosures in business disputes appears to have resulted in cases with fewer discovery motions and costs more proportionate to case type and the amount in controversy. Meanwhile in Utah, broad initial disclosure rules have seemingly led to quicker case dispositions, fewer discovery disputes in most types of cases, and, according to most attorneys, lower costs. Now years removed from the backlash against the 1993 amendments, many federal district courts have begun experimenting with requiring parties in certain employment disputes to provide certain disclosures automatically and early. And a study by the Federal Judicial Center shows that motions practice in these cases has fallen by over 40 percent.
Given all this evidence, it’s hard not to wonder if the real problem with the 1993 experiment was simply that it was ahead of its time. Maybe we just needed to wallow a little longer in collateral discovery disputes and watch them become ever more complicated and exasperating with the exponential growth of electronically stored information before we could appreciate this potential lifeline out. At the least, it would seem churlish to ignore all that’s happened since 1993 and not bother with a pilot project to test in the federal system more broadly what seems to be working so well in so many states and in a discrete set of cases in federal court.
The skyrocketing costs of legal education are no secret. Since the 1980s, private law school tuition in the United States has increased by 155.8 percent and public law school tuition by 428.2 percent (yes, in real, inflation-adjusted terms). Today, many students pay over $200,000 for a legal education — that on top of an equally swollen sum for an undergraduate degree. And with rising tuition costs come other costs too. Increased debt loads reduce students’ incentives and ability to take on lower-paying public service or “main street” legal jobs. No doubt, as well, some of these increased costs are ultimately borne by consumers, as lawyers pass along as much of their “overhead” expenses (student loans) as they can. Which raises the question: Why is a legal education so expensive?
It’s hard to ignore the possibility that our legal education accreditation requirements are at least partly to blame. Take California’s suggestive experience. In deference to the ABA, most states require anyone sitting for the bar to graduate first from an ABA-accredited law school. But in California it’s possible for graduates of state-accredited or unaccredited law schools to take the bar exam. And the cost differential is notable: average tuition runs $7,230 at unaccredited schools, $19,779 at California-accredited schools, and $44,170 at ABA-accredited schools in the state. No doubt the increased marketability of an ABA-accredited degree is responsible for some of the difference here. But isn’t it worth asking whether at least some of our often well-intended accreditation requirements are actually worth the costs they impose?
Consider first and perhaps most ambitiously the mandate that most everyone must attend three years of law school after the completion of a college degree. We’ve come a long way from Abraham Lincoln’s insistence that “[i]f you wish to be a lawyer, attach no consequence to the place you are in, or the person you are with; but get books, sit down anywhere, and go to reading for yourself. That will make a lawyer of you quicker than any other way.” For much of our nation’s history, President Lincoln’s advice held true: The only requirement to become a lawyer in most states was to pass the bar exam. Even some of the law’s luminaries as late as the mid-20th century didn’t attend three years of law school, greats like Justices Robert Jackson and Benjamin Cardozo and Harvard Law School Dean Roscoe Pound.
Where did the idea of three years of graduate education come from? It appears most states adopted the requirement at the behest of the ABA. In pushing states to adopt this requirement, the ABA emphasized that legal education must develop in students a mind attuned to the common law — an argument arguably not specific to three years as opposed, say, to two or four. The ABA also invoked the fact the American Medical Association had proposed a four-year standard for physicians and reasoned that, because law, like medicine, is a complex field, legal studies should last for a comparable period — an argument that seems to have stemmed more from professional pride than empirical proof.
Even if these doubtful rationales once seemed sufficient to persuade states to mandate a monolithic three-year graduate course of study, do they really remain persuasive today? Competitive and consumer-friendly markets are usually characterized by a diversity of goods, specialized to fit consumer needs and preferences — and markets with just one good of uniform character are often the product of a producer-friendly monopoly or some similar competitive failure. And while it would be wrong to suggest that all law school educations are identical, it might be worth asking whether three years (with a largely prescribed first year) is necessary for each and every law student. Is it truly the case that the legal training of a Main Street family lawyer needs to follow the same basic trajectory as a Wall Street securities lawyer, especially when demand for the former’s services is often acute and routinely unmet? Recently, the ABA acknowledged the need for greater heterogeneity in legal education. And one starting place might be to permit students to sit for the bar after only two years of study, allowing students and employers alike to determine the value of an optional third year of law school. President Obama, himself a Harvard-trained lawyer, has promoted this concept.
Consider that in the United Kingdom the legal education market is a good deal more heterogeneous than ours. To qualify for practice, a student may either take a three-year undergraduate course or a one-year graduate conversion course. Meanwhile, further graduate educational options are available in a variety of fields (e.g., criminal justice, intellectual property, and human rights) for those seeking specialized skills. But none of this is essential. After the basic academic instruction, a student may decide to become a barrister or solicitor. Depending on his or her choice, the student will then have to undertake additional training, often a one-year specialized educational course followed by a hands-on apprenticeship during which he or she will usually receive only modest compensation. But even the minimum wage presents a substantial swing from expending $50,000 or more on a year of formal legal education in the United States. This diversity of legal education options does not appear to be a threat to the rule of law in the United Kingdom — and it is difficult to see how it might be here.
Beyond that, we might also ask about the value of some of the more discrete accreditation requirements we impose on law schools today. In our zeal for high educational standards, we have developed a long and dreary bill of particulars every law school must satisfy to win ABA accreditation and it’s often unclear whether these many and various requirements can be justified on the basis of evidence of improved outcomes.
Here are just a few illustrations. Law schools must employ a full-time library director (dare not a part-timer) with the job security of a faculty position. And maybe that’s necessary after all because of some of the many other requirements that the ABA imposes on law school libraries — like the requirement they furnish a device to print microform documents. (Does anyone still use those? Or is there just one microfiche printer left, passed between law schools one step ahead of the accreditation committee?) Schools must extend extensive tenure guarantees to faculty, and full-time faculty must teach “substantially all” of a student’s first-year courses, even if adjuncts would prove just as good. Schools must also generally maintain student-faculty ratios of 30:1 or less (about the same ratio found in many public schools), though adjuncts (full disclosure: like me) count as only one-fifth of a professor for this purpose. Meanwhile, if professors have any sort of ongoing relationship with a law firm or business, a presumption arises that they are not full-time. And if an American law school wants to offer something other than a traditional JD program, like the sort of diverse degree programs found in English universities, it must receive a special dispensation from the ABA council responsible for legal education. Then, too, there are the restrictions on the number of credits a student may take at any given time, and the rule that no more than a third of credit hours can be earned for study or activity outside the United States. And beyond even that don’t forget that while students usually may receive credit for unpaid internships, they generally may not earn credit for the very same internship if it offers pay and helps reduce their debt load.
Of course, any revisions to our rules governing law schools would raise complicated cost-quality tradeoffs. Some believe that the current American legal education regime is necessary to permit future lawyers to develop sufficient knowledge of legal doctrine and capacity for legal analysis. Justice Antonin Scalia, for example, once argued that “the law-school-in-two years proposal rests on the premise that law school is — or ought to be — a trade school,” a premise he believed erroneous. Others defend the current system by citing familiar consumer-protection concerns. And others still point out that the third year offers opportunities to take elective courses in specialty areas of the law.
Admittedly, these seem good enough arguments to persuade a reasonable mind that at least some lawyers should undertake three years of graduate education. These also may be good enough arguments to justify imposing some significant restrictions on those who opt out of a third year (e.g., requiring on-the-job training for a period of years under the tutelage of a supervisor as in the English system). But it’s far less clear whether these are sufficient grounds for concluding that everyone needs three years of graduate legal training, or legal training shaped by so many and such detailed accreditation requirements. Commendably, a 2014 ABA whitepaper explored some of these questions and concluded that many current accreditation requirements do indeed increase cost without conferring commensurate educational benefits. As a result, the paper encouraged a shift from a regulatory scheme controlling so many detailed aspects of the educational process to a scheme focused more on outcomes and empirical cost-benefit analyses. And true to its word, the ABA’s section on legal education has begun relaxing at least some of its more extraordinary accreditation requirements. First steps, maybe, but steps in the right direction.
Lowering barriers to entry, ensuring judicial resolutions come more quickly and at less cost, and making legal education more affordable share the common aim of increasing the supply and lowering the price of legal services. All of these potential changes, too, are uniquely within our profession’s power to effect. Of course, meaningful change rarely comes easily, let alone when it requires a self-regulating profession to undertake self-sacrifice. But estimates suggest that inefficient policies and our professional regulations result in a roughly $10 billion annual “self-subsidy,” in the form of higher prices lawyers may charge their clients compared to what they could charge in a more competitive marketplace. Might not our willingness to confront candidly just how much of that self-subsidy is warranted prove a good test of our commitment to civil justice reform — and whether we as a profession wish to do good or merely do well?