Anomaly puzzle in N=1 supersymmetric electrodynamics as artifact of dimensional reduction
Abstract
Calculations of the two-loop
β
-function for N=1 supersymmetric electrodynamics are compared for regularizations by higher derivatives and by the dimensional reduction. The renormalized effective action are found to be the same for both regularizations. However, unlike the dimensional reduction, the higher derivative regularization does not lead to anomaly puzzle, because it allows to perform correct calculation of diagrams with insertions of counterterms. In particular, using this method a contribution of diagrams with insertions of counterterms is calculated exactly to all orders. This contribution appears to be 0 if the theory is regularized by the dimensional reduction. We argue, that this result follows from mathematical inconsistency of the dimensional reduction and is responsible for the anomaly puzzle.