Abstract: . . .  iso 14001 Page 1 Matthew Potoski Aseem Prakash Covenants with Weak Swords: ISO 14001 and Facilities’ Environmental Performance Journal of Policy Analysis and Management, Vol. 24, No. 4, 745–769 (2005) © 2005 by the Association for Public Policy Analysis and Management . . .
. . .  decisions. The relationship between the amount of time a facility was out of com- pliance in 1995 and 1996 and ISO 14001 registration follows an inverted U-shaped curve. The compliance coefficients are jointly significant, with the compliance 95–96 coefficient being positive and the compliance 1995–96 2 being negative. Facilities that are always in compliance or always out of compliance are the least likely to join ISO 21 An alternative . . .
. . .  treat- ment variable and uncorrelated with the errors of the outcome equation. Instrumental variable(s) that are weakly correlated with the treatment variable can be problematic (Bound, Jaeger, & Baker, 1995). A “weak” correlation is problematic when the joint F test of the instrumental variable(s) on the treat- ment (endogenous) variable is close to one (Bound et al., 1995, p. 446). F-tests on the instrumental vari- able(s) . . .
. . .  iso 14001 Page 1 Matthew Potoski Aseem Prakash Covenants with Weak Swords: ISO 14001 and Facilities’ Environmental Performance Journal of Policy Analysis and Management, Vol. 24, No. 4, 745–769 (2005) © 2005 by the Association for Public Policy Analysis and Management Published by . . .
. . .  jointly significant, with the compliance 95–96 coefficient being positive and the compliance 1995–96 2 being negative. Facilities that are always in compliance or always out of compliance are the least likely to join ISO 21 An alternative approach by Heckman and Robb (1985) for analyzing the impact of treatments on out- comes suggested is the first difference or fixed effects estimator. Taking advantage of panel data includ- ing . . .
. . .  significant at p 0.001. Facilities’ compliance histories have an important impact on their ISO 14001 decisions. The relationship between the amount of time a facility was out of com- pliance in 1995 and 1996 and ISO 14001 registration follows an inverted U-shaped curve. The compliance coefficients are jointly significant, with the compliance 95–96 coefficient being positive and the compliance 1995–96 2 being negative. Facilities . . .
. . .  advantage of panel data includ- ing at least one pre-treatment panel and one post-treatment panel, and assuming E(U it – U it 1 | d i , X it – X it 1 ,) 0 for all t, t 1, t k t 1, where U is the error term. Heckman and Robb (1985, p. 217) sug- gest the regression model (Y  . . .
--3000,7,214,3019,56539