Working Paper 277
Time to Build and Aggregate Fluctuations: Some New Evidence
Revised August 1, 1986
This paper presents maximum likelihood estimates of a real business cycle model very similar to one Kydland and Prescott  suggested. The results of the paper conflict with Kydland and Prescott’s. The model leaves unexplained much of the variance of two key investment series, namely, structures and equipment. Also, much of the variation in the differences of per capita hours can be generated assuming that past leisure choices do not affect current utility.
Download Paper (pdf)