Censoring, or zero expenditures, in the dependent variables of demand systems can lead to inconsistent parameter and elasticity estimates. Poi (2012) introduced the Stata command quaids to estimate quadratic almost-ideal demand systems, or QUAIDS (Banks, Bundell and Lewbel, 1997), although without the possibility to address censoring. In this paper, we introduce the command quaidsce to consistently estimate the QUAIDS in the presence of zero observations, using the two-step procedure proposed by Shonkwiler and Yen (1999). The new command also allows for estimating expenditure and price elasticities with standard errors, using post-estimation tools.
