Description

This study uses national accounts, household surveys, and administrative records to provide consistent distributional series, emphasizing labor income between 2006 and 2017. Our new methodology is able to correct the known limitations of different data sources and combine them coherently for the first time using Chilean data. In contrast to estimations in advanced economies we propose methods to estimate data that are not available, which could be use to estimate distributional series in other developing countries with limited official statistics. The validity of the imputations is verified by contrasting the results with various external references. On average, the underestimation of gross wages in the Chilean national household survey as compared to national accounts is 40%, significantly larger than other countries. About a quarter of this gap is attributed to the “missing rich” in the survey. For 2017, this equates to an estimated median gross income for dependent labor of CLP 600,000 and CLP 570,000 for all workers. The corrected mean-median income ratio (Gini) is 26% (17%) larger than in the raw survey of 2017, and falls only 6% (3%) between 2006 and 2017 compared with a larger decline of 12% (11%) in the original data.