Empirical issues in the computation of Stone-Lewbel price indexes in censored micro-level demand systems

In this article, we propose an empirical method for the computation of the Stone–Lewbel (SL) price index for product aggregates, when censored samples with zero expenditures are available from household budget surveys. The proposed technique is based on a regression imputation method that takes into account the price dynamics, therefore, allowing to disentangle the role of demographics from the role of prices in computing the SL index. Our simulations seem to indicate that our method is a valuable alternative.