Robert E. Marks
Australian Graduate School of Management,

Blame Beethoven. His Symphony No. 5 in C Minor, written in 1807, still requires as many people to play it as it did 200 years ago, and unless they play very fast, it will take about the same time too. That is, there has been no improvement in the players' productivity. Meanwhile, what with computers and multi-skilling, most of the rest of us have been improving ours, and the prices charged for our outputs have been falling, most notably in the IT area, not surprisingly. This mismatch is an example of Baumol's cost disease, cures for which are not obvious: culling orchestras and reducing repertoires might cut costs, but almost certainly they would cut revenues too. More paying customers would help, perhaps listening to duelling orchestras, or movie scores, as well as the classics. We see symptoms of the disease in other parts of the economy too: nurses, teachers, lawyers. Indeed, many labour-intensive industries in the service sector suffer.

In 1965 William Baumol distinguished between activities that could experience increases in output per employee ("productivity growth") from those that could not. Such increases might occur because of the increased used of tools, or machines, or other inputs per hour of labour, which means that his distinction is rooted in the nature of the production process or technology that produced the output, rather than in the nature of the output itself. Primary and secondary sector production fall into the first (increasing productivity) group, while services production by and large falls into the second (non-increasing productivity).

A brief diversion into economic theory: If an economy has two sectors, one in which output per labour hour is rising (increasing productivity) and one in which output per labour hour is constant (constant productivity), then the average cost of production in the first sector will be falling compared to the average cost of production of the second sector. The earnings of the employees in the first sector will rise faster than the earnings of the employees in the second sector since the former are more productive than the latter. Whether the demand for the output of the second sector will support the higher costs will depend on the demand for this output and on the degree of competition in the provision of its output. In general, employment in the second sector will grow compared to employment in the first sector. These were Baumol's insights: although employment would rise proportionately in the second sector, so would its cost of provision.

Throughout the industrialised world, the data show that employment in the services sector has been the fastest growing of all sectors. This growth is partly demand-driven -- for instance, society demanding more services -- but also partly because slower productivity growth in services will, with relative real output levels unchanged, cause employment to grow in services relative to other sectors, such as goods production, as Baumol's unbalanced growth model explains.

Another way of thinking about this: Consider activities such as university lecturing, live theatre, live symphonic concerts, hairdressing, and arguing for one's client in court. These are activities where it is difficult to reduce the hours of labour input by substituting, say, machines for people. As Baumol puts it, these are activities which have slow productivity growth. These are also activities whose costs are rising fastest, as their employees aspire to rising standards of living. In activities with faster productivity growth, employees' earnings can also rise, but substitution of machines for labour inputs means that the average costs of these activities may remain constant or fall even as their employees enjoy higher pay.

Unless services manage to find ways to improve their labour productivity, their average costs will continue to rise, even with no rise in their levels of activity, leading to issues such as that motivating this volume: the rising cost of legal services. A striking example of these rises in Australia has been premiums for medical malpractice insurance: it's not that doctors have been performing more negligently (although it may have been getting easier to find expert witnesses to testify against their professional colleagues), but that both a rising incidence of negligence suits and rising damages payouts have led to higher costs and so higher premiums.

For the symphony orchestras, if there is no way to reduce costs without also reducing quality (for instance by eliminating performance of the large romantic pieces), then perhaps, through audio or video recordings (not taping the Adelaide Ring Cycle is a prime lost opportunity) or through other ways to stimulate revenues by stimulating audience numbers, orchestras can continue to survive the Baumol cost disease for a while. Unfortunately, given continuing productivity gains elsewhere in the economy, it can't be cured.