The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Legal Productivity, the Cost Disease, and AI
It has been a while since my last post on the Volokh Conspiracy. In 2021, I became associate dean at George Washington and did not have time to write. Last year, I switched associate dean roles and my portfolio became smaller, so I was fortunate to have some time to return to scholarship and to complete several articles. I'll begin my return to blogging by writing a series of posts offering shorter versions of the key arguments in a recently completed article that I have now submitted to law reviews, entitled The Cost of Justice at the Dawn of AI.
The article explores how changes in the productivity of lawyers affect the legal system and how legal actors should prepare for a future that may feature lawyers who are either more or less productive than today, depending in part on how artificial intelligence develops. The article's simplest claim is that the legal literature ought to pay more attention to the productivity of the legal sector, because changes in legal productivity directly affect the cost of legal services. Legal costs in turn affect how successfully the legal system can perform the core mission of ensuring that like cases are treated alike while cases are treated differently when applicable legal principles so demand.
The preeminent economic model for considering changes in the costs of goods and services over time is William Baumol's cost disease. This model is perhaps best known as providing an explanation for why the costs of college education and of health care have increased faster than inflation in recent decades. The essential story is that those sectors have enjoyed lower productivity increases than productive sectors like agriculture, textiles, or technology, and so products have become more expensive. Arguments that any given market is afflicted with the cost disease can be contested. Maybe, one might argue, educational costs have risen in large part because the core educational product—a professor lecturing in a classroom—has been bundled with increasingly lavish complementary services, such as ever-improving food and ever-increasing numbers of administrators to help students navigate their way through college.
Whatever the case for any given market, however, the cost disease story is practically tautological. When markets enjoy technological advances that increase productivity per worker, costs decrease relative to the costs in industries where technology has been relatively stagnant. If costs in productive sectors are falling relative to costs in stagnant sectors, then costs in stagnant sectors must be rising relative to costs in productive sectors. If we thus control for the overall price level, which depends primarily on macroeconomic policy, relatively stagnant sectors must become relatively more expensive over time. We can debate in any given market, including the legal sector, whether stagnation is occurring, whether apparent relevant inflation is attributable to the cost disease or to improvements in quality. But the very definition of productivity implies that if an industry stagnates, its products will become more expensive than other industries.
The article considers the implications of legal productivity both retrospectively and prospectively. The retrospective question is whether lawyers' productivity has stagnated, and if so, how has that affected the legal system. The prospective question is whether we should expect artificial intelligence to increase the productivity of lawyers and thus reduce the costs of legal services relative to other goods and services in the economy. These inquiries frame the ultimate question of how legal actors might prepare for a world in which legal services might be considerably more or less expensive in relative terms than they are today.
These arguments will unfold in future blog posts (and of course are developed in full in the article). For now, I will note two points from different parts of the article that highlight how considering productivity changes, past and future, may matter for understanding the future of the legal system.
In the past few decades, lawyers and legal scholars have observed a great reduction in the number of cases brought to trial, in both civil and criminal courts. This transformation has been referred to as the "vanishing trial," and the name is not hyperbole. The proportion of federal cases ending in trial has fallen by more than half in the twenty years since academics began discussing the vanishing trial in earnest, and declines have also occurred in state courts and in other countries. John Langbein argued in the Yale Law Journal that the vanishing trial is attributable to changes in procedural rules, and while that may be a partial explanation, it fails to answer why trial rates have declined so systematically, rather than only at times of procedural reform. Although commentators certainly understand that the high cost of legal services helps explain why cases settle, they have not considered whether increases in costs over time, potentially attributable to the cost disease, may explain the historic trends. Vanishing trials are precisely what one would expect from simple models of settlement bargaining in an environment in which costs are increasing.
The prospective point is that if the legal system has developed based on implicit assumptions about the cost of legal services, changes in legal productivity may change the balance of power in legal regimes. For example, legislators may allow for longer maximum criminal sentences than the legislators would think justice requires, if the legislators think that long potential sentences are necessary to give prosecutors leverage in an environment in which it is not practical for prosecutors to bring many cases to trial. Suppose, however, that AI greatly reduces the time that pretrial work takes, for example by efficiently sifting through evidence, creating trial plans, and assisting in the drafting of pleadings. Armed with a credible threat to take more cases to trial, prosecutors should be expected to exact higher sentences in plea bargaining. Judges will disagree about whether such a change is salutary or pernicious, but awareness of how changes in productivity may revitalize or upset established legal regimes is the first step toward conscious consideration of how the legal system should evolve following productivity changes.
But has law really been stagnant? And if so, will AI reverse this trend? I'll address these questions in the next two blog posts.
