Do alternative measures of government result in alternative explanations for government size?

This note extends the work of Borcherding, Ferris and Garzoni (2003) on government size by considering how traditional tests respond to alternative definitions of government size. An error correction format is used to show that a) qualitatively all measures of size perform well, b) government consumption (plus transfers) works best when explaining short run (long run) changes and c) public choice and Kau/Rubin variables often perform differently with respect to the short and long run.