Is Total Expenditure a Valid Proxy for Income in Statistical Matching? A Counterfactual Experiment

A critical but very common experimental situation in statistical matching (SM), where an expenditure survey is linked to a living conditions survey that includes an estimate of income, is the absence within the common variables as a proxy of income or of consumption. Researchers have often been in the situation of giving up the matching and looking for data that already contained the characteristics they were looking for. The aim of this study is to test whether total expenditure is a statistically reliable common variable for income when matching two datasets that do not include the income variable or a proxy of it. Our initial results derived in terms of matching quality produce evidence in support of the plausibility of total expenditure as a substitute for income.