Five out of six private sector workers are employed in service-producing (rather than goods-producing) industries. Much of our domestic production is also in services. There is no doubt that computers, information technology and the internet have increased the productivity of service workers, especially those in business and scientific fields, over the past two decades. However these productivity gains are extremely difficult to measure. The Bureau of Labor Statistics makes a valiant effort to measure labor productivity gains in service industries. A more careful examination of the detailed productivity data released by the BLS at the end of May casts doubt on our ability to measure productivity in the service sector.
Measuring labor productivity gains in the goods producing sector is relatively straight-forward. An increase in labor productivity occurs when more physical units of output (of a given quality) are produced per hour worked. The biggest challenge in measuring productivity in goods-producing jobs is adjusting for changes in the quality of the good being produced.
In service-producing industries it is extremely difficult to disentangle increases in the cost of an hour of a service-producer’s time and the quantity of the services provided. Consider a situation where the cost of an hour of a lawyer’s time increased by 20% over a five-year period. Is that a price increase, holding the quantity and quality of services rendered? Or are more legal services being provided per hour because of productivity gains due to the internet and improvements in computer databases. The same question can be asked for engineers, architects, consultants or accountants.
The following chart illustrates BLS measures of productivity gains (in output per hour) in three service-producing industries between 1987 and 2010.
Productivity fell in bars (technically drinking places serving alcoholic beverages) over the past twenty-three years. It is unclear what makes a bartender less productive. Apparently Americans are sipping their alcoholic beverages more slowly than we were in 1987. It is even more surprising that janitorial service providers saw a bigger increase in productivity than engineering service providers over the past two decades. Again it is unclear whether the BLS can measure how much more productive janitors have become. Perhaps it takes 55.7% less time to clean floors than it did in 1987. Or maybe are floors aren’t quite as clean as they used to be.
It is imperative that we accurately measure productivity gains in the service economy when we try to measure real GDP growth and price inflation. If productivity measures for many detailed service-producing industries are inaccurate can more aggregated measures of real output and prices be reliable? It is a challenge to accurately measure the quality and quantity of output in many industries in our information and service-based economy. It isn’t exciting to devote more of our resources to developing better measures of prices and quantities in these industries, but it may be warranted.