Abstract
Bi-Directional Reiterative Truncated Projected Least Squares is used with quarterly Japanese data for 1955 to 2001 to find that the multiplier effect from increases in public and private investment was less than the multiplier effect from decreases in public and private investment between 1989 and 2001.
Acknowledgment
We appreciate the comments and suggestions of Charles Horioka and Kazuo Ogawa and the research assistance given by Marisca Meyers and Anubha Bansal.