Will the real elasticity of substitution 'in Norwegian dentistry' please stand up?

The non-homothetic production cost structure in Norwegian private dentistry involves more than two factor inputs. Consequently, this paper implements-separately for solo and group practices-three conceptually different measures of factor substitutions to infer the precise nature of input associations using translog cost model estimates based on 1993 data. We calculate own- and cross-price elasticities of factor demands, pairwise elasticities of substitution and their approximate S.E.s. We find that: (1) dentists and dental assistants cannot be consistently aggregated as one homogeneous labour input; (2) input demands are inelastic; (3) the theoretically restrictive Allen-Uzawa and the less restrictive shadow and Morishima elasticities of substitution are not equivalent; (4) dentists and auxiliary dental personnel relate as significant substitutes in solo practices and as significant complements in group practices; (5) 'supplies' (e.g., dental materials) in the aggregate are substitutes for the two types of dentistry labour; and (6) there appears to be a wider scope for factor substitutions in group rather than solo practices. Due to inelastic factor demands, opportunities for cost controls are limited despite some tendencies for factor interchange. Finally, future researchers of production costs involving more than two inputs should investigate the less restrictive, alternative measures of factor substitutions.