On the Sensitivity of Aggregate Productivity Growth Rates to Noisy Measurement

Aggregate rates of productivity growth are among the most closely watched indicators of economic performance. They are also among the most difficult to measure accurately. This paper explores the sensitivity of such rates to random measurement error using a simple generic model. The model allows for errors in the input and output components of the productivity ratio, with different variances, and for serial and cross correlation of the errors. The effects of the errors are considered from the point of view of growth rates themselves, changes in growth rates, and comparisons between rates in different countries.