Estimating saving functions in the presence of excessive‐zeros problems

Abstract

Summary Zero-inflated Tobit models and hurdle models are developed to estimate the determinants of saving in the form of voluntary contributions to personal pension plans (PPPs) and bank deposits in Britain. These two types of saving are affected by an excessive-zeros problem. We find that, especially for saving in the form of PPP contributions, the estimates derived from both the univariate and the bivariate zero-inflated Tobit models are consistent with the hypothesis that misreporting significantly contributes to the excessive-zeros problem. However, the former model provides more realistic estimates of the determinants of the two types of saving. The univariate and bivariate hurdle models, on the other hand, give a satisfactory explanation of the positive saving, but fail to explain the zero saving.