I've always been sheepishly interested in what "management consultants" at places like McKinsey do. I did not go down that path for many reasons, and I'm glad for that. But, conceptually, I enjoy the challenge of trying to comprehend a business or industry and usefully evaluate how external forces such as technology will shape it.
What is puzzling to me is that despite being a prestigious, white-collar profession, there is no industrial apparatus at a place like McKinsey who manufacture white papers and slide decks for the legal industry. There are certainly people who provide business consulting services to lawyers, but they often work individually and seem to fall backward into that job. The business school types also do not seem that interested. And some academics study law as business, but they are often social scientists.
Now that AI is in the air, there is tons of discourse about how AI will change the profession, but the content of that discourse exposes the industry's own lack of awareness about how it actually works. So over the next few posts, I'm going to write more systematically and try to lay the foundation for my own, admittedly idiosyncratic, business theory of law.
The Hegemony of Law & Economics
To the extent that business analysis of law exists, it draws heavily "law and economics," a half-century-old framework that is now the dominant mode of how professors teach law students to think about the law. In law and economics, legal rules should be economically "efficient," which could mean that a discrete outcome (a settlement or a contract) maximizes the economic welfare of both parties, or that a rule has some productive benefit to society as a whole. The economic analysis is both a shield—a long-standing rule might be explained, post hoc, on efficiency grounds—and a sword—if you do some math and show that a rule you don't like is "inefficient," the analysis stops there.
This thinking extends to how lawyers analyze law firms, and most articles by law professors and consultants about lawyer organizational behavior are written through an unabashedly economic lens. Indeed, one of the top legal consulting firms, quoted in this New York Times piece about the decisions by some law firms to cut deals with Trump, is literally called Adam Smith, Esq.
For a long time, really since attending law school, I've struggled with the primacy given to economics as a way of explaining social ("market") behavior in general, and law especially. To be clear, economics is certainly a useful and interesting way of understanding the world, but while I think it is a credible way for making sense of the price of oil futures or why nails cost as much as they do at the hardware store, I lose the thread when we try to apply it to as a universal discipline that can explain everything from crime to child-rearing. And while anyone is certainly allowed to take a model of how society behaves and apply it in new contexts, the mathematical turn in economics has led people who use an "economic" mode of thinking to believe they are applying a set of mathematical axioms rather than a theory of how society operates.
To offer an example, consider this post at Adam Smith, Esq.'s blog, about the Thomson Reuters annual "state of the legal market" report in 2024. The post unpacks the "puzzle" in this chart, which tells us that since before the 2008 recession, growth in law firm headcount is outpacing the total number of billable hours worked.
Setting aside major conceptual questions about what terms like "productivity" and "demand" mean in this context, the takeaway is that law firms are consistently acting "unproductively," they are hiring more lawyers ("supply"), than the market can bear. This is very aggravating to the author—"I can’t think of another substantial sector in the services economy whose behavior resembles this consistent and sustained pattern of growing capacity faster than demand"—and he seems optimistic that generative AI may effectively serve a market-correcting function. ("But what if superior financial out-performance is no longer 'nice work if you can get it,' but an existential requirement? Enter Generative A.I.")
Although notable, this is not a cherry-picked example from this blog or from law and business writing writ large. From a different angle, scholars of "access to justice" frequently use supply-and-demand rhetoric, arguing that people who want lawyers do not get them largely because of aggressive restrictions on supply. I don't actually disagree with this in principle: Economics is a helpful way of explaining some things. My concern more is that we're suffering from a crisis of intellectual hegemony where we lack the terms to explain anything that occurs "outside the model."
Efficiency in Practice
To explain what I mean, consider this 2006 article by Herbert Kritzer, a political scientist who made a living studying lawyers, about an insurance defense law firm. "Insurance defense" in this context means a law firm that primarily represents insurance companies defending claims made against policyholders, e.g., the at-fault driver in a car accident or a construction company that may be responsible for a construction defect. These policyholders typically have an agreement with the insurance company that allows the company to take over the defense of any claim against their policy, so that an insured individual who is at fault cannot go out and hire their own lawyer1. The insurance company picks the lawyer and employs an insurance adjuster to manage the litigation.
These types of cases are incredibly common, and insurance defense lawyers are ubiquitous. Just in my own observations, cases with an insurance company defendant constitute a plurality, if not a majority, of the standard civil case docket in Chicago. (It's worth mentioning that the other side of insurance defense is often a "settlement mill" attorney of the sort I described here.)
Kritzer was interested in insurance defense firms because their services are so ubiquitous and, in his words, "commoditized":
Insurance companies are large consumers of legal services. Given the quantity of such services they buy, these companies want what is akin to wholesale prices for those services unless they are confronting exceptional circumstances (i.e., the threat of a very large loss, probably eight figures or more). Because of this volume, insurance companies are able to secure hourly rates that are well below those that are paid by other commercial clients. Relatively few insurance defense lawyers in the Twin Cities are able to charge rates in excess of $160-$170 per hour. More typical rates are in the range of $110 to $120 per hour for associates and $140-$150 per hour for the most senior partners; at some firms, the rates charged by associates start out as low as $90 per hour. To put these rates in perspective, the going rate for auto mechanics in the Twin Cities at the time of my research was about $80 per hour, and for plumbers it was as much as $150 per hour.
To understand this world, Kritzer worked for 3.5 months as a paralegal at an insurance defense firm in Minneapolis, interviewing most of the attorneys along the way. In theory, this law firm should be a beacon of economic productivity.
And in a way, it is. The firm certainly has to compete for business; according to Kritzer, the purchasing power of insurance companies makes the company’s threats to hire a cheaper firm incredibly effective. The result, he writes, "is a variety of alternative fee arrangements"—no more billable hour!
But the day-to-day reality for lawyers resembles working the sort of dumb corporate job parodied in Office Space. Which is to say that while there is clearly a corporate incentive to act "efficiently," in practice everyone engages in compliance rituals to fulfill this corporate incentive that add up to something which appears quite inefficient.
For example, take "litigation budgets" required by insurance companies as an auditing measure, where the lawyer is required to file a document that discusses what expenses they plan to incur in a case. Although in theory this is an cost-control measure imposed on the fee-hungry lawyer, Kritzer writes that "[t]he more frequent problem that lawyers mentioned was simply getting the adjuster to respond to the request to proceed with some activity[.]" In practice, the litigation budget has some positive aspects (one lawyer mentions that doing the budget helps him plan out the case, which he might do otherwise) and some bad ones (another lawyer says that the insurance company wouldn't pay him for his work on a case that settled before he had time to submit a budget). Mostly, it seems like the lawyers develop individual workarounds that depend on the responsiveness or level of trust with the adjuster.
There are many such examples throughout the article, with the throughline being lawyers' complaints that their work is more heavily controlled than it had been 10-20 years ago, when the practice was, to use Kritzer's term, less commoditized. (As one lawyer says: "Much more control, or attempt to control the case, by the insurance companies, rather than to rely on my judgment.") Commenting on this, Kritzer writes that "[w]hile these kinds of changes are attributed to cost consciousness on the part of the insurance companies, they may also reflect consolidation within the insurance industry, and the problems that such consolidation create for management and control." In other words, in the commoditized law practice, the lawyers make a little less but end up dealing with significantly more administrative overhead than they did before. It's not clear that this is a good outcome for economic efficiency or human decency.
To be fair to the economists, I would agree that the commoditized insurance defense firm behaves, on balance, more "efficiently" than most firms. But setting aside whether that's good for insurance defense attorneys, it's not clear to me that at higher-end firms, clients want their attorneys to be this financially conscious. An important fact to remember about these insurance defense firms is that the insurance company that employs the adjuster is not the "client." That would be the person (the "insured") who holds the policy and committed whatever act that led to a lawsuit. So for the insurance company, these lawyers are truly more like a commodity, a cost of doing business that needs to be minimized.
But this is an exception. Like doctors, people and companies hire lawyers for all kinds of deeply personal and meaningful reasons that don't fit in the economist's model of the world, and it stands to reason that lawyers would manage their business differently (and perhaps less "productively") than a nail manufacturer. If we want to understand how legal practice operates—and thus how technologies like generative AI might affect this practice—we should be open to the idea that economics is not an intellectual Swiss Army knife for analyzing human behavior.
In need of a better model, let me suggest management cybernetics, or the study of control systems.
What the hell are you talking about?
"Cybernetics" as a field was invented by British and American scientists flush with capital and new ideas after World War II, who were interested in understanding the behavior of what they called "control systems," which you might call "robots." To do this, they needed a general mathematical theory of how systems receive and act on information from their environment.
The core concept of cybernetics is the law of "requisite variety," which says that every system adjusts to regulate the amount of information ("variety") it receives from its environment. It can do this through attenuation: Through the autonomic nervous system, your body constantly receives input from outside systems through nerves that it immediately discards. Or it can do it through amplification: A well-performing small business hires more people to manage its increased number of orders. Put differently, every system adapts to control its environment. The viable, self-regulating system thus becomes a "model" of its environment and survives by constantly adjusting its behavior in response to external feedback (the word "feedback" itself is a term coined by cybernetics).
A very simple model of a self-regulating system is a heating and air-conditioning system. It operates through a single regulator, the thermostat, which takes in feedback, the surrounding air temperature, and provides a simple binary instruction (on or off) to the heater or air conditioner. If the temperature is outside a certain threshold, the machine turns on until it passes the threshold, at which point it turns off again. And thus, the room stays at a specific temperature without any other intervention.
In business, the person who applied cybernetics to the study of management was a truly eccentric person named Stafford Beer, who I discovered through the writing of Dan Davies, a banker who recently published an excellent book, The Unaccountability Machine, that applies Beer's ideas to the failure of the modern economy.2 To oversimplify, Beer's big idea is that organizational charts and accounting tables, the classic tools of management are poor representations of what makes a business work. What matters for survival is the proper handling of information: Making sure that the right person has the information they need to make the right decision in the moment. As society and businesses become more larger complicated, handling information necessarily becomes more difficult, and organizations typically respond by trying to attenuate information.
Although he is not so bold as to call himself a management cybernetician, you can see these ideas at work in Kritzer's paper, particularly in the sentence that I italicized, where he recognizes that the insurance companies are responding to external changes (consolidation) by exerting much more "control" over cases. In the good old days, companies still exerted control over the case—they still employed adjusters to manage the lawyers—but with fewer overall claims the adjusters could work them on, literally, a case-by-case basis. As the number of claims grows the company adapts by making the handling of claims more routine through attenuation, reducing the complexity of the decisions made by individual adjusters. In the same way, before we had a kid, my spouse and I would get home at 5:30 and decide what to make for dinner. Now, we write out a meal plan every week and decide exactly what our hungry toddler will have in front of him by 6:00. As things get more complicated, i.e., as variety increases, systems adjust to stay regulated.
Beer's motto is that the "purpose of the system is what it does." That is, when presented with information about how something turned out, you should treat the inputs as a black box and focus on understanding the outcome. Returning to the chart where we started, instead of scolding law firms for not being "business-like," we should ask what it is about law firms, as a system, that consistently leads to a situation where they are hiring more employees relative to revenue growth than other industries. The law firms are surviving, and we should figure out why that is.
If you’re wondering how this is ethical—good question! The short answer is that insurance companies negotiated a carve-out to the general ethical principles for lawyers that instruct the lawyer to follow the client’s wishes above all others.
I wrote about Davies previously here. His blog, back of mind, is also a great read.