AI Can Cure Baumol’s Cost Disease - But Only If We Want It To
Image Source: Unsplash
Some sectors of the economy have costs that rise faster than in other sectors. This simple insight and its implications lead to concern about the cost of government services as well as the survival of the arts. Artificial intelligence is one tool that can help us overcome the problem.
The economist William Baumol is usually credited with the concept, often called Baumol’s Cost Disease or the Baumol Effect. He and a co-author wrote, “The output per man-hour of the violinist playing a Schubert quartet in a standard concert hall is relatively fixed, and it is fairly difficult to reduce the number of actors necessary for a performance of Henry IV, Part I.” In contrast, productivity in agriculture and manufacturing has hugely increased over many years. Wages tend to rise at the same pace as output per worker across the entire economy. Sectors with no or small productivity gains see rising labor costs without improvement in output per hour worked. Thus, Baumol argued, activities such as performing arts and teaching will become increasingly expensive.
Baumol’s insight was based on valid observations—but of the past. There is nothing inherent in manufacturing and services that make productivity gains different in one sector relative to the other. It seems to be an historical artifact of the technological progress of the time, not an immutable law of economics or technology.
Since Baumol’s work on the subject, retail trade productivity has increased substantially. He noted, back in the 1960s, that despite the development of supermarkets and better packaging, further gains in retailing productivity were unlikely. But today barcode scanners have eliminated the need to tag each product and for cashiers—when they are even used—to key in prices. Behind the scenes better inventory management has reduced unproductive effort and spoiled produce.
Scanning across different sectors of the economy, as McKinsey analysts have done, shows many service sectors doing well. The overall economy’s growth rate averages about one percent per year. But the sector “Arts, Entertainment and Recreation” surpassed the average gains in output per hour worked.
Consider Baumol’s comment about the Schubert concert. The violinist cannot play faster, but what about the marketing staff? Can tickets be sold more productively than through a person sitting in a booth? Can the sound and lighting be handled by a smaller staff than in years past?
Artificial intelligence can step in where people are performing intellectual tasks. Take education as an example. We often think of it as a teacher lecturing to a class, but consider all of the elements. The curriculum is developed, lesson plans are created, class progress is guessed, lesson plans are adjusted, students are evaluated. AI can help at each of those steps. Maybe a good AI tool can look at a group of students before the school year starts and adjust the standardized lesson plan for them. Maybe the AI can monitor each student’s learning day by day, then adjust future days’ lessons for the class. Maybe the AI can send a student computer-based lessons, games and quizzes that closely match that particular student’s needs and abilities. Maybe it can direct the teacher on group activities that will enhance the students’ learning. Perhaps one teacher can handle 40 or 50 or more students, with better results and less stress. I don’t know if this can happen, but it’s a vision that schools should attempt.
Some of the slow growth sectors are less competitive. A recent commentary applied Baumol’s idea to the latest British budget, explaining that there just isn’t much room for productivity growth in education or health care. One must ask how much these sectors have tried to improve productivity. Manufacturing companies can improve their profits by increasing productivity. If one firm lags behind their competitors in profitability, management will be replaced. The farmer who eschews improved techniques will earn a low return, so that land purchase offers by highly productive operators become very attractive.
In education, however, no principal tries to do a better job with fewer teachers. There’s no reward for it. In fact, administrators' compensation is more closely tied to the number of the people they supervise than the effectiveness of the schools.
The cure for Baumol’s cost disease will be application of all tools, including AI, to all challenges across the economy. The school principal and hospital administrator should be as diligent about finding productivity enhancements as the factory manager or farmer. In the public sector, that means political pressure on administrative leaders to find ways to do more with less, using whatever technology works best.
More By This Author:
The Fed And Federal Budget Extravagance: Interest Rate Cuts At RiskData Revisions Reveal: Consumers Have More Spending Power Than We Thought
The Rising Storm: Business Leaders Navigate America's Labor Shortage