Abstract
Experimental data on π± production by polarized photons are compared with a vector-dominance model prediction which is independent of a possible ambiguity regarding the frame used to evaluate the ρ0 density matrix; a discrepancy of more than a factor of 2 is found.