Long live the billable hour
Until we find something better
This is the next installment in my loosely connected "business of law" series, where I try to develop a clearer sense of how AI will actually affect legal practice. In the next two posts, I'm tackling the oft-discussed question of AI's impact on the billable hour. Most people with strong opinions on this topic usually rely on assumptions with a poor foundation, both about why lawyers use billable hours in the first place and what exactly it is about AI that will give them a reason not to.
I'll start with an examination of the billable hour, which has been a hobby horse of lawyers long before ChatGPT. Indeed, in 2002, ABA President Robert Hirshon wrote a sharp editorial in the ABA Journal, exhorting the bar "to start looking at value instead of cost when determining fair payment for services."
Seems sensible. But this bland call to arms begs the difficult question: How do you value a lawyer's services? Put differently, why do people hire lawyers for anything?
Let's take two examples. First, in response to the Trump administration's efforts to penalize law firms for DEI initiatives, David Lat has pointed out that many of the firms that caved to the administration instead of fighting the orders were more dependent on transactional partners whose (astronomical) value derives from their ability to close "deals." In this day and age, that means having the ear of the federal government, which is to say that the value of these partners is not about their technical expertise as much as it is about their relationships with the right people in the highest places.
We can also learn from looking further back in history to 1676, when George Cateret, proprietor of the American colony of East Jersey, granted land to Simon Rouse and his "Heires or assigns for ever." Rouse left the land by will to Frances Moore, and litigation arose in the 1690s over whether Moore had valid title. At issue was the fact that the land had not been left to Rouse and his "Heires and assigns." The difference in the single word meant that—according to the treatise of English Judge Sir Edward Coke—Rouse could not transfer his interest, and a jury ruled against Moore. As Lawrence Friedman writes in A History of American Law:
[T]he decision was a crucial one; it cast in doubt the rights of many landholders, who claimed through Moore. The province was thrown in an uproar. Remedial legislation was proposed, but the governor blocked it, hoping to exact payment quit-rents as the price of quieting the disputed titles. Ultimately, the proprietary party conceded the point, and an act was passed declaring that all of Governor Cateret's "grants, charters, or patents" bearing the "particle or...in the habendum" should be "taken, deemed and esteemed as effectual in law" as if the word and had appeared in place of the or.
Arguably, this case turns more on "legal skill" than those of the hypothetical Big Law dealmakers; in retrospect, Simon Rouse probably would have shelled out more for a lawyer who had read his Sir Edward Coke. But in context, the example speaks to the uncertain and contingent nature of law in the colonies, and Friedman's point is that no one really knew what rules should apply. In colonial America, as in the second Trump administration, the most valuable lawyers are prized for their ability to manufacture certainty in extremely uncertain times.
That sure sounds "valuable," but how do you charge for that?
The Billable Hour as Management Tool
Although Abraham Lincoln may1 have said "a lawyer's time and advice are his stock in trade," Lincoln, the practicing lawyer, did not bill by the hour and he did not manage a law firm. He had "partners," but, prior to the early 20th century, lawyers partnered only to save on overhead expenses; they did not share cases or clients.
The first "law firm," in the sense of a large legal practice owned by a partnership of attorneys with a staff of associates and other paraprofessionals, was Cravath, Swaine & Moore. Paul Cravath founded the firm on Wall Street in 1903 and developed what is known as the "Cravath system" for corporate law firms. In short, he hired associates from only the very best law schools and "trained" those associates into partners by working them to the bone for six years. Although the big law firms are much bigger and firms do not always follow the system to the letter, the spirit of the Cravath system very much lives on in corporate law, where associates are often expected to bill inhumane hours for exorbitant salaries.
But at the time of its founding, neither Cravath nor its peer firms on Wall Street, billed by the hour. Instead, they charged each client on a "value" basis, picking what seemed like a reasonable number for services rendered. The system worked because the firms openly colluded by agreeing not to steal each other's clients. And there was no competition for star lawyers. Once an associate was made partner, he was a partner for life—there was no chance of moving "laterally" to another firm.2
Instead, credit for invention of the billable hour goes to Reginald Heber Smith, a Boston attorney. Smith graduated from Harvard law in 1914 and went to work as general counsel of the Boston Legal Aid Society. Although his legacy today is controversial, Smith was a zealous advocate for legal aid in his time and wrote a widely read monograph in 1919, Justice and the Poor, the first systematic study of legal aid in the United States.
Smith was a devotee of "Taylorism," the theory of scientific management applied by Frederick Taylor to industrial manufacturing operations at the turn of the century. Because of failing eyesight, Taylor turned down admission to Harvard so he could work as an apprentice machinist in a factory; he quickly rose up the ranks and made his career by "scientifically" analyzing how to make factories more efficient. In a laudatory essay, Peter Drucker, an influential management theorist (who coined the term "knowledge work"), gives Taylor much of the credit for the American productivity revolution from the late 19th century to the middle of the 20th, through "the application of knowledge to work."
Smith applied these principles by tracking hours on timesheets at legal aid to clear cases more efficiently. In 1919 he moved to the Hale & Dorr law firm (which eventually merged with another firm and is now known as Wilmer Hale), where he served as managing partner for the remainder of his career. Smith patiently implemented his scientific management system at Hale & Dorr, and memorialized the system in Law Office Organization, a series of four articles published in the American Bar Association Journal in 1940.3
By that time, Smith had perfected the "daily time sheet," on which every Hale & Dorr attorney recorded each task in tenths of an hour, a division of time chosen "because it facilitates not only addition but other calculations...[f]or convenience in figuring nothing beats the decimal system."
Smith conceived of the time sheet only as an internal tracking system. The actual bills were determined at a weekly firm meeting, "an integral part of the plan of organization." Every partner and associate was invited, but not required, to attend, and the chairperson role was assigned on a rotating basis to every attorney, including associates.
Bills were a regular agenda item at the meeting. Although a partner could, with approval of another partner, send out a bill without waiting for firm meeting, the partners were encouraged to discuss the bill at the meeting first. Smith writes:
It is essential to the success of any law firm that its bills to its clients be fair and that they be satisfactory. The best guarantee is to submit a proposed bill to the judgment of other men. They do not necessarily fix the bill but out of their experience they can generally offer some closely parallel situations and tell what charge was made. The rule is that after hearing the opinions of his fellows, the responsible attorney is entitled to have the final say. He may have been moved upwards or downwards in his mind but at least he has been given all the help of which the firm is capable.
Although Smith had developed an elaborate cost accounting system—the description of which consumed an entire article in his four-part series—he wrote that "legal services, with few exceptions, cannot be standardized." Thus, while the report on what it cost to do the job "is an illuminating and steadying factor," he knew of no better method for exact determination than "open discussion with one's partners and associates." However, even after the bill was generated internally, the client was given the final say, and time records were provided upon request. If the firm felt the client was being unfair, it would simply refuse to accept further work from that client.
The 1958 Lawyer
Although, by all accounts, Smith's articles were immensely popular—the collection was sold as a pamphlet through the ABA—the next two decades were not kind to lawyers' wallets. Relative to other professions, such as doctors, lawyers' real incomes had stagnated. Although they were certainly not poor, there was great anxiety within the profession about the ability of lawyers to maintain the upper-middle-class lifestyle they felt entitled to.4
Law professors George B. Shepherd and Morgan Cloud argue in their 1999 article, Time and Money: Discovery Leads to Hourly Billing, that the drop in real incomes can be attributed to the profession's failure to switch from "value" billing to hourly billing after the expansion of pretrial discovery in the 1938 Federal Rules of Civil Procedure led to the drop in real incomes. Using a fancy economic model, they purport to show that the expansion of pretrial discovery increased the cost of litigation generally and increased uncertainty, in the sense that the cost of every individual case was now harder to predict. Although I find it hard to take the model seriously, I take their point that the expansion of pretrial discovery did lead to major changes in how litigation operated, and it's plausible that changes to the nature of the practice made it more difficult for lawyers to capture the value of their work using existing methods.
Whatever the cause, by the late 1950s, the American Bar Association decided it was time for lawyers to get their act together. A "Special Committee on Economics of Law Practice" was formed to "ascertain the causes which have resulted in the failure of lawyers to maintain an economic status comparable to that of persons in other professions, businesses and trades" and propose "remedial steps" for lawyers to recover their rightful economic status. The committee published a series of pamphlets, the first being The 1958 Lawyer and His 1938 Dollar. The committee took lawyers to task for their general failure at managing money, and suggested that "economical office operation" was one of the ways lawyers could increase their net earnings and "maintain their status" in society (re: keep up with doctors).
Channeling Smith, the committee recommended that lawyers start using a daily time sheet system, recording their work in tenths of an hour. Like Smith, the authors of the pamphlet do not suggest that the internal time records should reflect the final fee billed to the client. While the costs and time expended for the client determine in a "provisional way some basis for determining the fee to be charged," the ABA reminded lawyers of other "professional and economic factors" which should be considered when determining fees. These factors include some of the same basic factors considered by Smith and his colleagues at Hale & Dorr: The novelty and difficulty of the case, the skill required by the lawyer, and past dealings with the clients.
While this all sounds quite genteel, the pamphlet makes offhand reference to a more sinister motive for keeping time secret—the need to maintain minimum fee schedules. As discussed below, the Supreme Court had not yet outlawed these schedules, published by local bar associations to set a floor for what lawyers in the area would charge for certain routine services. The 1958 Lawyer warns lawyers that they would lose business if they used timesheet data to increase their charges above the local customary rates. Unmentioned is the alternative possibility: That radical billing transparency might have revealed to clients that they were being paying too much for legal services.
Social Shakeup - 1960s and 70s
Shortly after The 1958 Lawyer was published, demand for lawyers increased as federal and state legislatures enacted—and private citizens and governments enforced—a wide range of groundbreaking laws. Suddenly, companies had to navigate a world where, among other things, their employees had the right to sue them for race and sex discrimination, consumers had the right to sue for any deceptive or misleading conduct, and governments proactively regulated companies on a range of issues, including the environment, workplace safety, and product labeling.
The same spirit of the times5 led to changes to the laws regulating lawyers and their business practices. Three crucial legal developments affected the practice of law in the 1970s:
To encourage private enforcement, the newly passed federal laws allowed plaintiffs to recover attorneys' fees for successful lawsuits, but did not prescribe a method for calculating a fee award. In the early 1970s, federal appeals courts adopted the "lodestar" method of calculating attorneys' fees, which calculates a fee award based on the number of hours worked, multiplied by a "reasonable hourly rate" for the jurisdiction.
As mentioned above, in 1975, the Supreme Court held in Goldfarb v. Virginia State Bar that the legal profession was subject to antitrust laws, and therefore that the minimum-fee schedule for basic legal services was an illegal restraint of trade.
In 1977, the Court held in Bates v. State Bar of Arizona that lawyers had a First Amendment right to advertise their services.
As corporate legal needs grew, they began hiring professional "general counsel" to handle legal matters within the corporation, rather than outsourcing this work to "their" law firm, as had been the custom since the days of Cravath. As summarized in a 1985 article, these newly professionalized corporate counsel provided "preventive or anticipatory legal services, including longer range planning and programmatic prevention," and managed outside counsel.
Notably, it was the clients that demanded hourly billing, a move Herbert Kritzer attributed to the "accounting culture" within senior management:
[A]s the offices of corporate general counsel became professionalized (in a business sense, not a legal sense), there was a need to apply business principles to their operation, including the purchase of outside services. Hours and rates could be easily measured and compared, and general counsel began to demand that their outside law firms provide detailed bills with this information.
However, the new corporate-legal accounting paradigm failed to control costs. Instead, the push to bring legal work into the corporation correlated with an explosion in spending on legal services by corporations. Comparing data from a survey of Chicago lawyers in 1975 with a similar study conducted in 1995, John Heinz, Robert Nelson, and Edward Laumann found in 2001 that the number of lawyers and legal services consumed had increased in absolute terms, and that most of the growth was accounted for by more lawyers doing corporate legal work.
Rise of Alternative Fees
By the late 1980s and 1990s, it had become clear that law firms were not managing their end of the bargain, using hourly bills as a blank check to engage in billing practices that ranged from questionable to grossly unethical. In response, corporate clients began pushing for so-called "alternative fee" arrangements, and the pages of law reviews and lawyer trade magazines spilled over with stories of fed-up corporate counsel finally taking control of their legal spending.
For example, the "Dupont model," created in the 1990s by general counsel for the eponymous corporation, was widely adopted by companies in the late 1990s and early 2000s. As described in a 2004 ABA Journal article, the model involved using far fewer law firms (Dupont downsized from 350 outside firms to 41), then developing a "close relationship" and "detailed playbook" with the remaining firms to measure best practices. Similarly, a 2012 ABA Journal article describes Tyco International Inc.'s effort to bring down legal costs in products liability cases by imposing an alternative fee system called the "convergence" model. Essentially, instead of paying various law firms by the hour on a case-by-case basis, Tyco initiated a bidding process whereby it agreed to give all of its products liability cases to one firm, for a flat fee with guardrails (e.g., bonuses and penalties for coming in over or under budget).
It is not a coincidence that models like Dupont's and Tyco's arose with the emergence of the Internet, which enabled the company to (literally) invite law firms into its network and monitor them more closely. Better technology, in turn, enabled further monitoring. In that 2004 article, Thomas Sager, Dupont's general counsel and the originator of the model, looks ahead to the next frontier—digital billing records that are "suitable for crunching." As Sager told the author:
"So there's been a significant shift in control in how we review, what's covered and what's kicked back," says Sager. "Now we're in control like never before."
And in the article about Tyco, the author notes that firms using alternative billing are only able to do so with more sophisticated accounting models.
Standard billing and tracking models can't provide the analysis necessary for lawyers to effectively and accurately predict how a future matter should be priced[.] Firms need to develop the infrastructure and fee-analysis programs to effectively implement alternative pricing models from cradle to grave, including initial proposals, case management, staffing standards, revenue analysis, and post-engagement reviews.
Today, many firms of course still bill by the hour, but these stories make clear that for at least 30 years, corporate clients and law firms are behaving (for the most part) like sophisticated businesses. Moreover, the idea of using technology to better "value" a case is clearly not new, but it does not change the fact that legal services are inherently difficult to price. Generative AI's impact is better understood as part of this ongoing give-and-take, rather than a deus ex machina that will pull corporate counsel and law firms out of the dark ages.
NB: He probably didn’t say this.
This account is drawn from the book White Shoe, about the rise of Wall Street firms in the early 1900s.
I have not found an entirely convincing explanation for this decline. Labor statistics show a marked increase in the number of law school attendees after World War II, which led to a subsequent oversupply of lawyers, but that does not explain the relative decline in incomes—presumably medical schools were filled with former soldiers, too. One possible explanation is that, around this time, medical schools began capping enrollment at the behest of the American Medical Association (these caps remain in place today). Although many lawyers advocated for similar enrollment caps in law schools, those restrictions were never actually implemented.
If you’re interested in this time period, I highly recommend Public Citizens: The Attack on Big Government and the Remaking of American Liberalism. It’s a concise history about Ralph Nader and the rise of the nonprofit legal advocacy movement he championed in the 1970s. The Goldfarb and Bates decisions are directly attributable to the efforts of Nader and others.