How model dependent are sparticle mass bounds from CERN LEP?

In order to translate nonobservation of events of a type not expected in the standard model into bounds on particle masses, one necessarily has to make some specific assumptions. We systematically study the model dependence of the bounds that can be derived from the negative outcome of sparticle and Higgs-boson searches at CERN LEP as well as from the measurement of the total and nonhadronic decay widths of the $Z$ boson. We also show that the constraints from direct searches taken together with those from the width measurements imply stronger bounds than those obtained using either strategy by itself. Regarding model dependence, we find that the bounds on charged-sparticle and sneutrino masses that have been obtained with the usual assumptions for their decay patterns are, in fact, generally valid within the framework of the minimal supersymmetric model (MSSM). In models with a conserved $R$ parity, these bounds can be significantly reduced only by the ad hoc introduction of two charged SU(2)-singlet and/or an additional pair of SU(2)-doublet superfields. The bounds on neutralino masses, on the other hand, are very sensitive to even minor changes of the minimal model. At present, even in the MSSM the mass of the lightest neutralino is only mildly constrained and, in fact, cannot be bounded in models with just a slightly altered neutralino sector unless a supersymmetry signal is found. We also comment on the implications of a possible negative outcome of sparticle searches in future experiments.