On Top of the Top: A Generalized Approach to the Estimation of Wealth Distributions. Disslbacher, F., Ertl, M., List, E., Mokre, P., & Schnetzer, M. July 2023. Unpublished manuscript
On Top of the Top: A Generalized Approach to the Estimation of Wealth Distributions [link]Link  doi  abstract   bibtex   6 downloads  
The wealth distribution is infamously top-heavy, while the decisive upper tail is missing from survey data on household wealth in European countries. We provide a novel quantile regression approach to estimate all parameters of the Pareto and Generalized Pareto distribution to adjust for the rich missing in survey data due to differential non-response and under-reporting. In contrast to existing Pareto-based adjustment routines, the generalized and rules-based method is scalable, flexible in the face of heterogeneities in data quality and wealth accumulation regimes, transparent, and prevents over-shooting of wealth aggregates and wealth concentration estimates. We apply the method to data on fourteen Eurozone countries by supplementing the Household Finance and Consumption Survey (HFCS) with a novel database on country-specific rich lists from the European Rich List Database (ERLDB) compiled from country-specific rich lists. The magnitude of the resulting upper-tail adjustments varies substantially across countries, highlighting the importance of the rules-based method developed here. In addition, while the results are highly stable across an extensive range of sensitivity tests addressing the opacities of ERLDB data, the resulting estimates vary substantially across parameters borrowed from prior work.
@unpublished{Disslbacheretal2023,
  title = {On Top of the Top: A Generalized Approach to the Estimation of Wealth Distributions},
  author = {Disslbacher, Franziska and Ertl, Michael and List, Emanuel and Mokre, Patrick and Schnetzer, Matthias},
  year = {2023},
  month = jul,
  doi = {10.2139/ssrn.4499915},
  url = {https://doi.org/10.2139/ssrn.4499915},
  abstract = {The wealth distribution is infamously top-heavy, while the decisive upper tail is missing from survey data on household wealth in European countries. We provide a novel quantile regression approach to estimate all parameters of the Pareto and Generalized Pareto distribution to adjust for the rich missing in survey data due to differential non-response and under-reporting. In contrast to existing Pareto-based adjustment routines, the generalized and rules-based method is scalable, flexible in the face of heterogeneities in data quality and wealth accumulation regimes, transparent, and prevents over-shooting of wealth aggregates and wealth concentration estimates. We apply the method to data on fourteen Eurozone countries by supplementing the Household Finance and Consumption Survey (HFCS) with a novel database on country-specific rich lists from the European Rich List Database (ERLDB) compiled from country-specific rich lists. The magnitude of the resulting upper-tail adjustments varies substantially across countries, highlighting the importance of the rules-based method developed here. In addition, while the results are highly stable across an extensive range of sensitivity tests addressing the opacities of ERLDB data, the resulting estimates vary substantially across parameters borrowed from prior work.},
  keywords = {Cross-National Comparisons,Trends in Aggregate Wealth and Wealth Inequality},
  note = {Unpublished manuscript}
}

Downloads: 6