Abstract
This paper studies households’ response behaviour in a wealth survey. We analyse how unit non-response and item non-response contribute to the estimated distribution of wealth. Our findings imply that wealth inequality is underestimated in the survey. The downward bias is originating from item non-response and not from unit non-response. Wealthier households are less likely to provide answers to wealth-related questions. As a result, the level of net wealth is underestimated and the top tail of its distribution is missing. Imputation can eliminate biases throughout most of the wealth distribution but does not recover the estimates in the top tail.
This is a preview of subscription content, log in via an institution to check access.
Source: Calculations of the authors from the Estonian HFCS
Source: Calculations of the authors from the Estonian HFCS
Similar content being viewed by others
Notes
Examples of wealth surveys are the Bank of Italy’s SHIW survey, the Federal Reserve Board’s SCF, the Bank of Spain’s EFF and the ECB’s HFCS.
There are actually only four income components that need to be imputed, because one of the income components, private pensions, had no missing observations. See “Appendix 1” Table 2 for the related statistics.
The following description of methods proceeds from the description of the commands “mi impute chained” and “mi impute pmm” in the Stata manual.
As imputed values are drawn from the observed data, this method also preserves the distribution of the observed data. We have tested alternative imputation methods such as the linear regression, but this gives a very similar distribution of the imputed data to that from predictive mean matching.
This weak link between net wealth and unit response is not caused by controlling for income deciles in the regressions, as the link is still weak when the income deciles are excluded from the regression.
The Estonian HFCS aims at oversampling the richest households. The oversampling is based on personal incomes of the survey contact people in the period prior to the survey (the income data are obtained from the registers). To test the sensitivity of our estimates to oversampling, we re-estimated the results shown in Tables 2 and 4 removing the oversampled contact persons randomly. This had no significant effect on the estimation results, and therefore the estimates based on this exercise are not reported.
References
Alvaredo F, Saez E (2010) Income and wealth concentration in Spain in a historical and fiscal perspective. In: Atkinson AB, Piketty T (eds) Top incomes global perspective. Oxford University Press, Oxford
Bach S, Thiemann A, Zucco A (2018) Looking for the missing rich: Tracing the top tail of the wealth distribution. DIW discussion paper no. 1717
Bover O (2011) The Spanish survey of household finance (EFF): description and methods of the 2008 wave. Documentos Ocasionales no 1103, Banco De España
Bricker J, Henriques A, Krimmel J, Sabelhaus J (2016) Measuring income and wealth at the top using administrative and survey data. Brook Pap Econ Act 2016:261–312
Brzezinzki M, Salach K, Wronski M (2020) Wealth inequality in Central and Eastern Europe: evidence from household survey and rich lists’ data combined. Econ Transit Inst Change 28:637–660
Chakraborty R, Waltl SR (2018) Missing the wealthy in the HFCS: micro problems with macro implications. European Central Bank, working paper series, 2163
Cowell FA, Van Kerm P (2015) Wealth inequality: A survey. J Econ Surv 29(4):671–710
D’Alessio G, Faiella I (2002) Non-response behaviour in the Bank of Italy’s Survey of Household Income and Wealth. Banca D’Italia working paper no 462
Dell F, Piketty T, Saez E (2007) Income and wealth concentration in Switzerland over the 20th century. In: Atkinson AB, Piketty T (eds) Top incomes over the 20th century. Oxford University Press, Oxford
Gelman A, Hill J (2006) Data analysis using regression and multilevel/hierarchical models. Columbia University Press, New York
HFCS (2013) The Eurosystem household finance and consumption survey: results from the first wave. ECB statistics paper series no 2
HFCS (2017) The household finance and consumption survey: methodological report for the second wave. ECB statistics paper series no 17
Johansson F, Klevmarken A (2007) Comparing register and survey wealth data. Department of Economics, Uppsala University, Uppsala
Kennickell AB, Woodburn RL (1999) Consistent weight design for the 1989 and 1995 SCFs, and the distribution of wealth. Rev Income Wealth 45(2):193–215
Lundberg J, Waldenström D (2018) Wealth inequality in Sweden: what can we learn from capitalized income tax data? Rev Income Wealth 64(3):517–541
Meriküll J, Rõõm T (2019) Estonian household finance and consumption survey: results from the 2017 wave. Bank of Estonia occasional paper no 1/2019
Neri A, Ranalli MG (2012) To misreport or not to report? The measurement of household financial wealth. Banca D’Italia working paper no 870
Osier G (2016) Unit non-response in household wealth surveys. ECB statistics paper series no 15
Pérez-Duarte S, Sánchez-Muñoz C, Törmälehto V-M (2010) Re-weighting to reduce unit non-response bias in household wealth surveys: a cross-country comparative perspective illustrated by a case study. In: Proceedings of the European conference on quality in official statistics
Piketty T (2014) Capital in the twenty-first century. The Belknap Press of Harvard University Press, Cambridge
Roine J, Waldenström D (2009) Wealth concentration over the path of development: Sweden, 1873–2006. Scand J Econ 111(1):151–187
Roine J, Waldenström D (2015) Long-run trends in the distribution of income and wealth. In: Atkinson AB, Bourgiognon F (eds) Handbook of income distribution, vol 2. North-Holland, Amsterdam
Vermeulen P (2016) Estimating the top tail of the wealth distribution. Am Econ Rev Pap Proc 106(5):646–650
Vermeulen P (2018) How fat is the top tail of the wealth distribution? Rev Income Wealth 64(2):357–387
Yan T, Curtin R (2010) The relation between unit nonresponse and item nonresponse: a response continuum perspective. Int J Public Opin Res 22(4):535–551
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Authors declare that they have no conflict of interest and the research has been conducted without involving animal participants and the data of human participants has been treated by meeting confidentiality standards. The views expressed are those of the authors and do not necessarily represent the official views of the Bank of Estonia or the Eurosystem.
Additional information
The authors would like to thank the participants at the Bank of Estonia research seminar, at the HFCN meeting in Krakow and at the Joint Statistical Meetings conference in Denver, Colorado for their insightful comments.
Appendices
Appendix 1: Descriptive statistics of unit and item response rates
Appendix 2
See Table 8.
Rights and permissions
About this article
Cite this article
Meriküll, J., Rõõm, T. Are survey data underestimating the inequality of wealth?. Empir Econ 62, 339–374 (2022). https://doi.org/10.1007/s00181-021-02030-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-021-02030-6
Keywords
- Wealth distribution
- Unit non-response
- Item non-response
- Participation bias
- Wealth survey
- Wealth inequality
- Household Finance and Consumption Survey
- Estonia