AGGREGATION AND ACCURACY IN MEASURING TOTAL FACTOR PRODUCTIVITY: EVIDENCE FROM RAIL PRODUCTIVITY STUDIES

This paper provides empirical evidence on the magnitude of errors in measuring total factor productivity which can be introduced by use of aggregate data. The empirical cases are for the rail industry, and use three separate data sets (U.S. railroad, Canadian railways, and CP Rail). For all three cases, the magnitude of bias resulting from use of aggregate output data (ton-miles) is provided, and for one case, data were available on the bias from using aggregate labor data. The results show that significant bias is present: from 0.7 percent to 1.4 percent per year for use of aggregate rail output data. All three cases result in an overstatement of output and hence productivity growth. For use of aggregate labor data, the bias was 0.2 percent, also resulting in an overstatement of true productivity growth, compounding the errors from output aggregation.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00719250
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Mar 27 1996 12:00AM