The applicability of self-reported home values in housing wealth inequality assessment: evidence from an emerging country

Abstract

One of the measures of housing wealth inequality is the property’s market value. In existing analyses, this figure is often a subjective value determined by homeowners. Little is known about the validity of using this type of data as a substitute for market value in inequality studies. Therefore, this paper aims to examine whether self-reported home values can be applied to evaluate housing wealth inequality. In order to achieve this goal, first, a theoretical framework on the irrelevance of valuation bias for the assessment of housing wealth inequality was developed, followed by an empirical analysis. The latter included gathering information on subjective flat values and their characteristics in Warsaw. Next, a geographically weighted regression was calibrated to calculate the market value of these dwellings. Then, using the Gini coefficient, housing wealth inequality levels were estimated separately for subjective and objective home values. The results revealed that the former could serve as a very good proxy for the latter in housing wealth inequality analysis. The findings hold across almost all identified subgroups based on respondents’ gender, age, income, wealth, education and employment status. Finally, recommendations were formulated for public institutions and housing research.

Keywords:

• Geographically weighted regression
• housing wealth inequality
• Warsaw housing market
• subjective and objective measures

Acknowledgements

I would like to thank Martin Lux and three anonymous reviewers for their valuable comments.

Declaration of interest statement

No potential conflict of interest was reported by the author.

Notes

1 The Polish Register of Prices and Values of Real Estate serves only to collect property transaction prices. The collection of data on property values was terminated on 1 September 2017 when a provision requiring Polish valuers to submit extracts of their property value opinions to the Register of Prices and Values of Real Estate was repealed.

2 Research on misperceptions of inequality has been undertaken, but this relates to income inequality (Hauser & Norton, 2017).

3 Dummy variables were not logarithmic.

4 Due to the fact that this survey lacks information on the exact locations of the dwellings of the interviewed respondents, it is not possible to account for the spatial autocorrelation phenomenon more accurately.

5 Please note that index i refers to dwellings from the survey questionnaire database, and index j to dwellings from the transaction price database.

6 The disadvantage of variance is that it is a measure whose values are unbounded; consequently, it is impossible to define a maximum difference between estimates that, if not exceeded, would entitle one to conclude that the subjective values are a very good, good, or at least fair proxy for objective ones.

Additional information

Funding

This research is supported by the National Science Centre, Poland under Grant 2020/04/X/HS4/00458. MT is also supported by the Foundation for Polish Science (FNP) under the START program.