In a recent issue this Review, Frank de Leeuw and Nkanta Ekanem (L-E) proposed to determine whether the marginal cost housing is constant or rising as a function of the amount housing services provided in an area.? I propose to show that a nonrising supply curve is inconsistent with positive long-run land prices and to show that the L-E housing supply elasticities, calculated from reduced form equations, are very sensitive to the specification the true supply and demand functions. An alternative and perhaps more reasonable a priori specification is herein set out which is then used to calculate much different elasticities supply. These alternatively calculated coefficients are elastic, rather than inelastic as deduced by L-E, and consistent with estimates which I have made in another paper, and with the reduced form estimated by the authors. The L-E model and estimates are shown to suffer from misspecification, statistical inconsistency, and multicollinearity.