In 1966, the economists William Baumol and William Bowen published an influential book, Performing Arts: The Economic Dilemma. The book described how certain labor-intensive industries had minimal productivity gains, yet showed continued increases in wages. It used the example of a modern day string quartet playing Beethoven for a live audience. Four musicians were required for the performance; no different from the 1700’s. However, wages for musicians rose at a rate greater than their growth in productivity. Baumol and Bowen posited that this was due to the need to compete with high productivity industries that could afford to pay higher wages. This idea came to be known as Baumol’s cost disease and is used to explain wage inflation in health care, education and government. A nurse can only change one bandage at a time and a professor can only teach one particular class.
In the 1960’s, most of the improvements in productivity came from capital investments in machinery. Today, we are living in an age where information technology is the primary driver of improved business processes. It’s a phenomenon that the MIT researcher Andrew McAfee refers to as digitization. An unprecedented confluence of improvements in mobile devices, broadband connectivity and social media is profoundly benefiting enterprises. But despite this, do we still see evidence of Baumol’s diseases around us?
In the current day, those functions or processes that don’t leverage digital efficiency gains represent a new “cost disease”, however, a curable one. They seem to come in two flavors: those processes constrained by laws of physics and those limited by laws of biology. Let’s start with physics. Just like Baumol’s string quartet, the movement of people and physical objects has seen limited efficiency gains. Since the invention of commercial jet airliners, there has been no material decrease in flight times. Looking forward, hypersonic jets are predicted to radically improve the speed of air travel. However, most experts believe it will be decades before we see the promised 5 to 10 times increases in speed. Similar constraints are seen in other areas of travel from cars to trains to ships. This means that business processes built around such physical processes as transportation (e.g. commuting, business travel, shipping) will not see material improvements. In contrast, substitute digital processes exist in the form of remote access, virtual meeting spaces, and video conferencing. Each of these digital processes has seen dramatic efficiency gains (i.e. functionality to price ratio) in recent years. And while the physical world processes stagnate, the digital processes should continue to widen their efficiency advantage.
In regards to biology, I am referring to the inherent limitations we have as human beings. We have limited attention spans, limited memory and finite endurance. Like our pre-digital predecessors, we still need 6-8 hours sleep. As with airline travel, our human abilities are unlikely to advance quickly. These biological limitations make us poorly suited for tasks involving repetitive precision, unerring consistency and endurance. Therefore, such tasks as monitoring or complex assembly/installation of critical systems will disfavor human involvement. Other examples of human constrained processes in the administrative realm are provisioning, order processing and logistics. Our ability to perform these tasks will be limited in terms of efficiency or accuracy improvements. While there have been technical solutions (e.g. ERP, automated monitoring, automated configuration tools) for decades, many firms still cling to manual effort for the processes listed above. Instead, BPM tools can act as a backstop, serving as a source of automation for those remaining processes that have no other solution or still need a “bridge” to the manual world.
So while there are still industries and functions with the properties of Baumol’s disease, technology provides options to reduce their impact. Organizations need to start by identifying the processes that will inherently lag in productivity gains. They should then consider two courses of action:
- Automation – Use technology to eliminate the inefficient process through a substitute capability
- Lean management methods – Eliminate non-value adding steps from the process
While both of these courses of action would seem obvious, organizational culture can act as a brake, slowing their implementation. Those firms with strong histories of bureaucracy, conservatism and tradition will show strong resistance to these changes. A common argument against elimination of manual processes is around ROI. If the payback is not clear and quick, it will be argued that the effort and risk is not worthwhile. These arguments typically underweight the rapid growth in technical capability that will continue to widen the efficiency advantage over physical processes over time. Those organizations that fail to acknowledge and eradicate cost disease will wind up at an operating disadvantage to startups or more progressive legacy competitors.
So was Baumol correct? Are there still limitations where certain industries must rely on manual effort? In my opinion, these limitations exist only in a very strict sense. It does still require 4 musicians to play a string quartet to a live audience. However, streaming video and HDTV radically expand the size of the audience that can watch the event with near-live quality. A nurse can only change one bandage at a time. However, telemedicine, remote patient monitoring, and service robots can all make the nurse more efficient. A professor is still limited to teaching a single class at a time. However online learning dramatically expands the class size.
Baumol’s mistake was not accounting for the dramatic impact that technology would have over virtually every business process in the decades following his seminal book. His thoughts still provide an important cautionary note for managers: The world of manual processes is lawfully constrained to show minimal improvements over time.