Threshold effects in non-dynamic panels: Estimation, testing, and inference

Threshold regression methods are developed for non-dynamic panels with indi vidual-specific fixed effects. Least squares estimation of the threshold an d regression slopes is proposed using fixed-effects transformations. A non- standard, asymptotic theory of inference is developed which allows construc tion of confidence intervals and testing of hypotheses. The methods are app lied to a 15-year sample of 565 US firms to test whether financial constrai nts affect investment decisions. (C) 1999 Elsevier Science S.A. All rights reserved.